A direct formulation for sparse (2009)
Michael I. Jordan, Laurent El Ghaoui
PCA using semidefinite programming
Spectral Clustering with Perturbed Data (2009)
Ling Huang, Donghui Yan, Michael I. Jordan, Nina Taft
Spectral clustering is useful for a wide-ranging set of applications in areas such as biological data analysis, image processing and data mining. However, the computational and/or communication...
DiscLDA: Discriminative Learning for Dimensionality Reduction and Classification (2009)
Simon Lacoste-julien, Fei Sha, Michael I. Jordan
Probabilistic topic models have become popular as methods for dimensionality reduction in collections of text documents or images. These models are usually treated as generative models and trained...
Probabilistic Inference in Queueing Networks (2009)
Charles Sutton, Michael I. Jordan
Although queueing models have long been used to model the performance of computer systems, they are out of favor with practitioners, because they have a reputation for requiring unrealistic...
Spectral Clustering with Perturbed Data (2009)
Ling Huang, Donghui Yan, Michael I. Jordan, Nina Taft
Spectral clustering is useful for a wide-ranging set of applications in areas such as biological data analysis, image processing and data mining. However, the computational and/or communication...
Nonparametric Bayesian Learning of Switching Linear Dynamical Systems (2009)
Emily B. Fox, Erik B. Sudderth, Michael I. Jordan, Alan S. Willsky
Many nonlinear dynamical phenomena can be effectively modeled by a system that switches among a set of conditionally linear dynamical modes. We consider two such models: the switching linear...
Kurt T. Miller, Thomas L. Griffiths, Michael I. Jordan
Nonparametric Bayesian models are often based on the assumption that the objects being modeled are exchangeable. While appropriate in some applications (e.g., bag-ofwords models for documents),...
High-dimensional union support recovery in multivariate (2009)
Guillaume Obozinski, Martin J. Wainwright, Michael I. Jordan
regression
Shared Segmentation of Natural Scenes UsingDependentPitman-Yor Processes (2009)
Erikb Sudderth, Michael I. Jordan
We develop a statisticalframework forthe simultaneous, unsupervised segmentation and discovery of visual object categories from image databases. Examining a large set of manually segmented scenes, we...
Efficient Inference in Phylogenetic InDel Trees (2009)
Alexandre Bouchard-côté, Michael I. Jordan, Dan Klein
Accurate and efficient inference in evolutionary trees is a central problem in computational biology. While classical treatments have made unrealistic site independence assumptions, ignoring...
Posterior Consistency of the Silverman g-prior in Bayesian Model Choice (2009)
Zhihua Zhang, Michael I. Jordan, Dit-yan Yeung
Kernel supervised learning methods can be unified by utilizing the tools from regularization theory. The duality between regularization and prior leads to interpreting regularization methods in terms...
Kernel dimension reduction in regression (2009)
Fukumizu, Kenji, Bach, Francis R., Jordan, Michael I.
We present a new methodology for sufficient dimension reduction (SDR). Our methodology derives directly from the formulation of SDR in terms of the conditional independence of the covariate $X$ from...
DiscLDA: Discriminative Learning for Dimensionality Reduction and Classification (2009)
Simon Lacoste-julien, Fei Sha, Michael I. Jordan
Probabilistic topic models have become popular as methods for dimensionality reduction in collections of text documents or images. These models are usually treated as generative models and trained...
The Sticky HDP-HMM: Bayesian Nonparametric Hidden Markov Models with Persistent States (2009)
Fox, Emily B., Sudderth, Erik B., Jordan, Michael I., Willsky, Alan S.
We consider the problem of speaker diarization, the problem of segmenting an audio recording of a meeting into temporal segments corresponding to individual speakers. The problem is rendered...
Ben Taskar, Michael I. Jordan, Simon Lacoste-julien
We present a simple and scalable algorithm for large-margin estimation of structured models, including an important class of Markov networks and combinatorial models. The estimation problem can be...
Percy Liang, Dan Klein, Michael I. Jordan
The learning of probabilistic models with many hidden variables and nondecomposable dependencies is an important and challenging problem. In contrast to traditional approaches based on approximate...
Gad Kimmel, Michael I. Jordan, Eran Halperin, Ron Shamir, Richard M. Karp
Population stratification can be a serious obstacle in the analysis of genomewide association studies. We propose a method for evaluating the significance of association scores in whole-genome...
Joint covariate selection and joint subspace selection for multiple classification problems (2009)
Obozinski, Guillaume, Taskar, Ben, Jordan, Michael I
We address the problem of recovering a common set of covariates that are relevant simultaneously to several classification problems. By penalizing the sum of ℓ2-norms of the blocks of...
Yin, Junming, Jordan, Michael I., Song, Yun S.
Motivation: Two known types of meiotic recombination are crossovers and gene conversions. Although they leave behind different footprints in the genome, it is a challenging task to tease apart their...
Xuanlong Nguyen, Martin J. Wainwright, Michael I. Jordan
We establish a general correspondence between two classes of statistical functions: Ali-Silvey distances (also known as f-divergences) and surrogate loss functions. Ali-Silvey distances play an...
Communicated by Jeffrey Elman Adaptive Mixtures of Local- Experts (2008)
Terence D. Sanger, Robert A. Jacobs, Michael I. Jordan, Steven J. Nowlan, Geoffrey E. Hinton, Cnrladn Ms Ia
receptor fields. yorks of locally-tuned nalysis of survey data, or approximation and riable interpolation: A and M. G. Cox, eds., 3ooks, New York. Learning internal rep-
Francis R. Bach, Michael I. Jordan
We present a class of algorithms for learning the structure of graphical models from data. The algorithms are based on a measure known as the kernel generalized variance (KGV), which essentially...
Laurent El Ghaoui, Michael I. Jordan
The minimax probability machine (MPM) considers a binary classification problem, where mean and covariance matrix of each class are assumed to be known. Without making any further distributional...
On surrogate loss functions and f-divergences (2008)
Xuanlong Nguyen, Martin J. Wainwright, Michael I. Jordan
The goal in the binary classification problem is to estimate a discriminant function γ from observations of covariate vectors and corresponding binary labels. We consider an elaboration of this...
Ben Taskar, Michael I. Jordan, Simon Lacoste-julien
We present a simple and scalable algorithm for large-margin estimation of structured models, including an important class of Markov networks and combinatorial models. The estimation problem can be...
Matthias Seeger, Michael I. Jordan
The favourable scaling behaviour of sparse approximations to Bayesian inference for Gaussian Process models makes them attractive for large-scale applications. We show how to generalize the...
DiscLDA: Discriminative Learning for Dimensionality Reduction and Classification (2008)
Lacoste-Julien, Simon, Sha, Fei, Jordan, Michael I.
Probabilistic topic models have become popular as methods for dimensionality reduction in collections of text documents or images. These models are usually treated as generative models and trained...
Estimating divergence functionals and the likelihood ratio by convex risk minimization (2008)
Nguyen, XuanLong, Wainwright, Martin J., Jordan, Michael I.
We develop and analyze $M$-estimation methods for divergence functionals and the likelihood ratios of two probability distributions. Our method is based on a non-asymptotic variational...
RUNNING HEAD: Probabilistic inference in graphical models Correspondence: (2008)
A “graphical model ” is a type of probabilistic network that has roots in several different research communities, including artificial intelligence (Pearl, 1988), statistics (Lauritzen, 1996),...
On the Inference of Ancestries in Admixed Populations (2008)
Sriram Sankararaman, Gad Kimmel, Eran Halperin, Michael I. Jordan
Inference of ancestral information in recently admixed populations, in which every individual is composed of a mixed ancestry (e.g., African Americans in the US), is a challenging problem. Several...
Conditionally Trained Latent Dirichlet Allocation for Text Modeling and Categorization (2008)
Simon Lacoste-julien, Fei Sha, Michael I. Jordan
The hierarchical modeling paradigm of parametric and nonparametric Bayesian statistics have found a widespread application in modeling data in many areas, for instance, text and language processing,...
Xuanlong Nguyen, Martin J. Wainwright, Michael I. Jordan
optimal quantization rules for some
Kenji Fukumizu, Francis R. Bach, Michael I. Jordan
We propose a novel method of dimensionality reduction for supervised learning. Given a regression or classification problem in which we wish to predict a variable Y from an explanatory vector X, we...
Union support recovery in high-dimensional multivariate regression (2008)
Obozinski, Guillaume, Wainwright, Martin J., Jordan, Michael I.
In the problem of multivariate regression, a K-dimensional response vector is regressed upon a common set of p covariates, with a p by K matrix B* of regression coefficients. We study the behavior of...
Consistent probabilistic outputs for protein function prediction (2008)
Obozinski, Guillaume, Lanckriet, Gert, Grant, Charles, Jordan, Michael I, Noble, William
Abstract In predicting hierarchical protein function annotations, such as terms in the Gene Ontology (GO), the simplest approach makes predictions for each term independently. However, this approach...
Peña-Castillo, Lourdes, Tasan, Murat, Myers, Chad L, Lee, Hyunju, Joshi, Trupti, Zhang, Chao, ...
Abstract Background: Several years after sequencing the human genome and the mouse genome, much remains to be discovered about the functions of most human and mouse genes. Computational prediction of...
Michael I. Jordan, Daniel M. Wolpert
We discuss some of the computational approaches that have been developed in the area of motor control. We focus on problems relating to motor planning, internal models, state estimation, motor...
Alice X. Zheng, Ben Liblit, Michael I. Jordan, Alex Aiken
We present a novel strategy for automatically debugging programs given sampled data from thousands of actual user runs. Our goal is to pinpoint those features that are most correlated with crashes....
Global Temperature Data Image Data (2008)
Jens Nilsson, Fei Sha, Michael I. Jordan
A methodology for discovering a data manifold that best preserves information relevant to a nonlinear regression. Regression on Manifolds using Kernel Dimension Reduction
RUNNING HEAD: Probabilistic inference in graphical models Correspondence: (2008)
Michael I. Jordan, Yair Weiss, Michael I. Jordan
A “graphical model ” is a type of probabilistic network that has roots in several different research communities, including artificial intelligence (Pearl, 1988), statistics (Lauritzen, 1996),...
AKernel-Based Learning Approach to Ad Hoc Sensor Network Localization (2008)
Xuanlong Nguyen, Michael I. Jordan, Bruno Sinopoli
We show that the coarse-grained and fine-grained localization problems for ad hoc sensor networks can be posed and solved as a pattern recognition problem using kernel methods from statistical...
SHADOWER: A generalized hidden Markov phylogeny for multiple-sequence (2008)
Jon D. Mcauliffe, Lior Pachter, Michael I. Jordan
functional annotation
Michael I. Jordan, Bruno Sinopoli
kernel-based learning approach to ad hoc sensor
Satinder P. Singh, Tommi Jaakkola, Michael I. Jordan
It is widely accepted that the use of more compact representations than lookup tables is crucial to scaling reinforcement learning(RL) algorithms to real-world problems. Unfortunately almost all of...
ABSTRACT Bug Isolation via Remote Program Sampling ∗ (2008)
Ben Liblit, Alice X. Zheng, Alex Aiken, Michael I. Jordan
We propose a low-overhead sampling infrastructure for gathering information from the executions experienced by a program’s user community. Several example applications illustrate ways to use...
Jinwen Ma, Lei Xu, Michael I. Jordan
It is well known that the convergence rate of the expectation-maximization (EM) algorithm can be faster than those of convention �rst-order iterative algorithms when the overlap in the given...
Ling Huang, Anthony Joseph, Nina Taft, Michael I. Jordan
We consider the problem of network anomaly detection in large distributed systems. In this setting, Principal Component Analysis (PCA) has been proposed as a method for discovering anomalies by...
Andrew Y. Ng, Michael I. Jordan
We present a class of approximate inference algorithms for graphical models of the QMR-DT type. We give convergence rates for these algorithms and for the Jaakkola and Jordan (1999) algorithm, and...
Alice X. Zheng, Alex Aiken, Michael I. Jordan
Abstract As part of our work on Cooperative Bug Isolation (CBI)we have undertaken to instrument and distribute binaries for a number of large open source projects. This public de-ployment is an...
A Permutation-Augmented Sampler for DP Mixture Models (2008)
Percy Liang, Michael I. Jordan, Ben Taskar
We introduce a new inference algorithm for Dirichlet process mixture models. While Gibbs sampling and variational methods focus on local moves, the new algorithm makes more global moves. This is done...
Neil D. Lawrence, Michael I. Jordan
We present a probabilistic approach to learning a Gaussian Process classifier in the presence of unlabeled data. Our approach involves a “null category noise model ” (NCNM) inspired by ordered...
Martin J. Wainwright, Michael I. Jordan
by penalized convex risk minimization
Joint covariate selection for grouped classification (2008)
We address the problem of recovering a common set of covariates that are relevant simultaneously to several classification problems. We propose a joint measure of complexity for the group of problems...
Xuanlong Nguyen, Martin J. Wainwright, Michael I. Jordan
optimal quantization rules for some
AKernel-Based Learning Approach to Ad Hoc Sensor Network Localization (2008)
Xuanlong Nguyen, Michael I. Jordan, Bruno Sinopoli
We show that the coarse-grained and fine-grained localization problems for ad hoc sensor networks can be posed and solved as a pattern recognition problem using kernel methods from statistical...
Discussion of boosting papers (2008)
Peter L. Bartlett, Michael I. Jordan, Jon D. Mcauliffe
The authors have contributed three significant papers that provide, among other insights, an understanding of the consistency of several “large margin ” methods for pattern classification. In...
Martin J. Wainwright, Michael I. Jordan
by penalized convex risk minimization
Percy Liang, Dan Klein, Michael I. Jordan
The learning of probabilistic models with many hidden variables and nondecomposable dependencies is an important and challenging problem. In contrast to traditional approaches based on approximate...
Francis R. Bach, Michael I. Jordan
Spectral clustering refers to a class of techniques which rely on the eigenstructure of a similarity matrix to partition points into disjoint clusters with points in the same cluster having high...
ABSTRACT Bug Isolation via Remote Program Sampling ∗ (2008)
Ben Liblit, Alice X. Zheng, Alex Aiken, Michael I. Jordan
We propose a low-overhead sampling infrastructure for gathering information from the executions experienced by a program’s user community. Several example applications illustrate ways to use...
Emanuel Todorov, Michael I. Jordan
1. Optimal control of modified Linear-Quadratic-Gaussian (LQG) systems All simulations described in the main text are instances of the following general model: where the it,
Francis R. Bach, Michael I. Jordan
We present a class of algorithms for learning the structure of graphical models from data. The algorithms are based on a measure known as the kernel generalized variance (KGV), which essentially...
In Neural Computation, 3, pages 79-87. Adaptive Mixtures of Local Experts (2008)
Robert A. Jacobs, Michael I. Jordan, Steven J. Nowlan, Geoffrey E. Hinton
We present a new supervised learning procedure for systems composed of many separate networks, each of which learns to handle a subset of the complete set of training cases. The new procedure can be...
Kenji Fukumizu, Francis R. Bach, Michael I. Jordan
We propose a novel method of dimensionality reduction for supervised learning. Given a regression or classification problem in which we wish to predict a variable Y from an explanatory vector X, we...
A randomization test for controlling population stratification in (2008)
Gad Kimmel, Michael I. Jordan, Eran Halperin, Ron Shamir, Richard M
whole-genome association studies
Emanuel Todorov, Michael I. Jordan
Behavioral goals are achieved reliably and repeatedly with movements rarely reproducible in their detail. Here we offer an explanation: we show that not only are variability and goal achievement...
Approximate Inference Algorithms for Two-Layer (2008)
Bayesian Networks Andrew, Andrew Y. Ng, Michael I. Jordan
We present a class of approximate inference algorithms for graphical models of the QMR-DT type. We give convergence rates for these algorithms and for the Jaakkola and Jordan (1999) algorithm, and...
A Kernel-Based Learning Approach to Ad Hoc Sensor (2008)
Network Localization Xuanlong, Michael I. Jordan, Bruno Sinopoli
this paper we propose a method that bypasses the ranging step altogether. We show that it is possible to pose a coarse-grained localization problem as a discriminative classification problem that can...
Discussion of Boosting Papers (2008)
Peter Bartlett Bartlett, Michael I. Jordan, Jon D. Mcauliffe
oosing f from the function class k (H, #) = {f # H : #f#H # #} so as to minimize R # (f ). . The paper by Lugosi and Vayatis considers the loss function #(#) = exp(-#), the function class b (G, #) =...
On Optimal Quantization Rules for Sequential (2008)
Decision Problems Xuanlong, Xuanlong Nguyen, Martin J. Wainwright, Michael I. Jordan
We consider the problem of sequential decentralized detection, a problem that entails the choice of a stopping rule (specifying the sample size), a global decision function (a choice between two...
Structured Prediction via the Extragradient (2008)
Method Ben Taskar, Ben Taskar, Michael I. Jordan, Simon Lacoste-julien
We present a simple and scalable algorithm for large-margin estimation of structured models, including an important class of Markov networks and combinatorial models. We formulate the estimation...
Modular And Hierarchical, Michael I. Jordan, Robert A. Jacobs, Michael I. Jordan
this article we discuss the problem of learning in modular and hierarchical systems. Modular and hierarchical systems allow complex learning problems to be solved by dividing the problem into a set...
Learning from Dyadic Data (2008)
To Appear In, Thomas Hofmann, Jan Puzicha, Michael I. Jordan
Dyadic data refers to a domain with two #nite sets of objects in which observations are made for dyads, i.e., pairs with one element from either set. This type of data arises naturally in many...
Structured Prediction via the Extragradient (2008)
Method Ben Taskar, Ben Taskar, Michael I. Jordan, Simon Lacoste-julien
We present a simple and scalable algorithm for large-margin estimation of structured models, including an important class of Markov networks and combinatorial models. We formulate the estimation...
To appear: Statistical Science, Special Issue on Bayesian Statistics. (2008)
Graphical Models Michael, Michael I. Jordan
this article our principal focus is on the presentation of graphical models that have proved useful in applied domains, and on ways in which the formalism encourages the exploration of extensions of...
Probabilistic Inference in Graphical Models (2008)
Jordan Cs Berkeley, Yair Weiss, Michael I. Jordan, Michael I. Jordan
this article has arisen through several di#erent historical strands. We briefly summarize these strands here and note some of the linkages with developments in the neural network field
[Abstract] [Full Text] [PDF] Movement Smoothness Changes during Stroke Recovery (2008)
Emanuel Todorov, Michael I. Jordan, J Neurophysiol, T. E. Hudson, L. T. Maloney, J. A. Goble, ...
You might find this additional information useful... This article cites 25 articles, 5 of which you can access free at:
Multiway Spectral Clustering: A Margin-based Perspective (2008)
Zhihua Zhang, Michael I. Jordan
Spectral clustering is a broad class of clustering procedures in which an intractable combinatorial optimization formulation of clustering is “relaxed ” into a tractable eigenvector problem, and...
Hierarchical Bayesian Nonparametric Models with Applications ∗ (2008)
Yee Whye Teh, Michael I. Jordan
Hierarchical modeling is a fundamental concept in Bayesian statistics. The basic idea is that parameters are endowed with distributions which may themselves introduce new parameters, and this...
Union support recovery in high-dimensional multivariate (2008)
Guillaume Obozinski, Martin J. Wainwright, Michael I. Jordan
regression
On the inference of ancestries in admixed populations (2008)
Sankararaman, Sriram, Kimmel, Gad, Halperin, Eran, Jordan, Michael I.
Inference of ancestral information in recently admixed populations, in which every individual is composed of a mixed ancestry (e.g., African Americans in the United States), is a challenging problem....
Tommi S. Jaakkola, Michael I. Jordan
Graphical models provide a formalism in which to express and manipulate conditional independence statements. Inference algorithms for graphical models exploit these independence statements, using...
LETTER Communicated by Radford Neal Attractor Dynamics in Feedforward Neural Networks (2007)
Lawrence K. Saul, Michael I. Jordan
We study the probabilistic generative models parameterized by feedforward neural networks. An attractor dynamics for probabilistic inference in these models is derived from a mean field approximation...
Michael I. Jordan, Zoubin Ghahramani, Tommi Jaakkola, Marina Meila, Lawrence Saul
ffl A wide variety of models are being studied in the machine learning literature, including:-- neural networks-- decision trees-- variants of mixture models-- variants of hidden Markov models--...
Convergence study and improvement of variational methods with MCMC (2007)
Nando De Freitas, Pedro Højen-Sørensen, Michael I. Jordan, Stuart Russell, De Freitas, Michael I
In this paper, we apply Markov chain Monte Carlo methods to improve the accuracy and modeling exibility of variational techniques. We use variational schemes to obtain an initial approximation to the...
2003/09/29 09:33 1 Kernel-based Integration of Genomic Data using Semidefinite Programming (2007)
Nello Cristianini, Michael I. Jordan, William Stafford Noble
An important challenge in bioinformatics is to leverage different descriptions of the same data set, each capturing different aspects of the data. Many such sources of information [about genes and...
Semidefinite relaxations for approximate inference on (2007)
Martin J. Wainwright, Martin J. Wainwright, Michael I. Jordan, Michael I. Jordan
graphs with cycles
Michael I. Jordan, Nello Cristianini, Nello Cristianini, Laurent El Ghaoui, ...
Learning the Kernel Matrix with Semi-Definite Programming
Abstract--The Expectation-Maximization (EM) algorithm is an iterative approach to maximum likelihood parameter estimation. Jordan and Jacobs recently proposed an EM algorithm for the mixture of...
MIT Center for Cognitive Science Occasional Paper 40 Forward models: Supervised learning (2007)
Michael I. Jordan, David E. Rumelhart
Internal models of the environment have an important role to play in adaptive systems in general and are of particular importance for the supervised learning paradigm. In this paper we demonstrate...
C.B.C.L. Paper No. 130 Factorial Hidden Markov Models (2007)
Zoubin Ghahramani, Michael I. Jordan
This publication can be retrieved by anonymous ftp to publications.ai.mit.edu. We present a framework for learning in hidden Markov models with distributed state representations. Within this...
David A. Cohn, Zoubin Ghahramani, Michael I. Jordan
This publication can be retrieved by anonymous ftp to publications.ai.mit.edu. For manytypes of learners one can compute the statistically \optimal " way to select data. We review how these...
C.B.C.L. Paper No. 108 Learning from incomplete data (2007)
Zoubin Ghahramani, Michael I. Jordan
This publication can be retrieved by anonymous ftp to publications.ai.mit.edu. Real-world learning tasks often involve high-dimensional data sets with complex patterns of missing features. In this...
Andrew Y. Ng, Michael I. Jordan
We compare discriminative and generative learning as typied by logistic regression and naive Bayes. We show, contrary to a widelyheld belief that discriminative classiers are almost always to be...
Andrew Y. Ng, Michael I. Jordan
We compare discriminative and generative learning as typied by logistic regression and naive Bayes. We show, contrary to a widelyheld belief that discriminative classiers are almost always to be...
Jinwen Ma, Lei Xu, Michael I. Jordan
It is well known that the convergence rate of the expectation-maximization (EM) algorithm can be faster than those of convention �rst-order iterative algorithms when the overlap in the given...
Michael I. Jordan, David E. Rumelhart
Internal models of the environment have an important role to play in adaptive systems in general and are of particular importance for the supervised learning paradigm. In this paper we demonstrate...
De Freitas, Michael I. Jordan, Stuart Russell
We propose a new class of learning algorithms that combines variational approximation and Markov chain Monte Carlo (MCMC) simulation. Naive algorithms that use the variational approximation as...
Tommi S. Jaakkola, Michael I. Jordan
variational approach to Bayesian logistic regression models and
Tommi Jaakkola, Michael I. Jordan, Satinder P. Singh
1 Recent developments in the area of reinforcement learning have yielded a number of new algorithms for the prediction and control of Markovian environments. These algorithms, including the TD()...
Presidential Young Investigator. (2007)
Convergence results for mixtures of experts 1 Convergence results for the EM approach to mixtures of experts architectures The Expectation-Maximization (EM) algorithm is an iterative approach to...
Martin Wainwright And, Martin J. Wainwright, Martin J. Wainwright, Michael I. Jordan, Michael I. Jordan
We present a new method for calculating approximate marginals for probability distributions defined by graphs with cycles, based on a Gaussian entropy bound combined with a semidefinite outer bound...
Sensorimotor Adaptation of Speech I: (2007)
Compensation And Adaptation, John F. Houde, Michael I. Jordan
When motor actions (e.g., reaching with your hand) adapt to altered sensory feedback (e.g., viewing a shifted image of your hand through a prism), the phenomenon is called sensorimotor adaptation...
Graphical models: Probabilistic inference (2007)
Jordan Cs Berkeley, Yair Weiss, Michael I. Jordan, Michael I. Jordan
this article has arisen through several dierent historical strands. We briey summarize these strands here and note some of the linkages with developments in the neural network eld
Statistical Debugging in the Presence of Multiple Errors (2007)
Ben Liblit Mayur, Ben Liblit, Mayur Naik, Alice X. Zheng, Alex Aiken, Michael I. Jordan
We present a statistical debugging algorithm that operates on very sparsely sampled data drawn from large numbers of user runs. By identifying program behaviors that significantly increase the...
The nested Chinese restaurant process and Bayesian inference of topic hierarchies (2007)
Blei, David M., Griffiths, Thomas L., Jordan, Michael I.
We present the nested Chinese restaurant process (nCRP), a stochastic process which assigns probability distributions to infinitely-deep, infinitely-branching trees. We show how this stochastic...
Hierarchical beta processes and the Indian buffet process. This volume (2007)
Romain Thibaux, Michael I. Jordan
We show that the beta process is the de Finetti mixing distribution underlying the Indian buffet process of [2]. This result shows that the beta process plays the role for the Indian buffet process...
The infinite PCFG using hierarchical Dirichlet processes (2007)
Percy Liang, Slav Petrov, Michael I. Jordan, Dan Klein
We present a nonparametric Bayesian model of tree structures based on the hierarchical Dirichlet process (HDP). Our HDP-PCFG model allows the complexity of the grammar to grow as more training data...
Hierarchical beta processes and the Indian buffet process. This volume (2007)
Romain Thibaux, Michael I. Jordan
We show that the beta process is the de Finetti mixing distribution underlying the Indian buffet process of [2]. This result shows that the beta process plays the role for the Indian buffet process...
Nonparametric estimation of the likelihood ratio and divergence functionals (2007)
Xuanlong Nguyen, Martin J. Wainwright, Michael I. Jordan
divergence functionals
Xuanlong Nguyen, Martin J. Wainwright, Michael I. Jordan
divergence functionals and the likelihood ratio by
Nonparametric estimation of the likelihood ratio and divergence functionals (2007)
Xuanlong Nguyen, Martin J. Wainwright, Michael I. Jordan
divergence functionals
The infinite PCFG using hierarchical Dirichlet processes (2007)
Percy Liang, Slav Petrov, Michael I. Jordan, Dan Klein
We present a nonparametric Bayesian model of tree structures based on the hierarchical Dirichlet process (HDP). Our HDP-PCFG model allows the complexity of the grammar to grow as more training data...
The infinite PCFG using hierarchical Dirichlet processes (2007)
Percy Liang, Slav Petrov, Michael I. Jordan, Dan Klein
We present a nonparametric Bayesian model of tree structures based on the hierarchical Dirichlet process (HDP). Our HDP-PCFG model allows the complexity of the grammar to grow as more training data...
Xuanlong Nguyen, Martin J. Wainwright, Michael I. Jordan
by convex risk minimization
Comment on "Support Vector Machines with Applications" (2006)
Bartlett, Peter L., Jordan, Michael I., McAuliffe, Jon D.
Comment on "Support Vector Machines with Applications" [math.ST/0612817]
On optimal quantization rules for some problems in sequential decentralized detection (2006)
Nguyen, XuanLong, Wainwright, Martin J., Jordan, Michael I.
We consider the design of systems for sequential decentralized detection, a problem that entails several interdependent choices: the choice of a stopping rule (specifying the sample size), a global...
Structured Prediction, Dual Extragradient and Bregman Projections (2006)
Taskar, Ben, Lacoste-Julien, Simon, Jordan, Michael I.
We present a simple and scalable algorithm for maximum-margin estimation of structured output models, including an important class of Markov networks and combinatorial models. We formulate the...
Convergence Results for the EM Approach to Mixtures of Experts Architectures (2006)
The Expectation-Maximization (EM) algorithm is an iterative approach to maximum likelihood parameter estimation. Jordan and Jacobs (1993) recently proposed an EM algorithm for the mixture of experts...
On the Convergence of Stochastic Iterative Dynamic Programming Algorithms (2006)
Jaakkola, Tommi, Jordan, Michael I., Singh, Satinder P.
Recent developments in the area of reinforcement learning have yielded a number of new algorithms for the prediction and control of Markovian environments. These algorithms, including the TD lambda)...
Strategic and Tactical Decision-Making Under Uncertainty (2006)
Jordan, Michael I., Anantharam, Venkat, El Ghaoui, Laurent, Russell, Stuart, Sastry, Shankar, Koller, Daphne, ...
This report presents the final conclusions of the research on decision-making under uncertainty conducted by the investigators at the University of California at Berkeley, Stanford University, and...
Learning spectral clustering, with application to speech separation (2006)
Francis R. Bach, Michael I. Jordan, Yoshua Bengio
Spectral clustering refers to a class of techniques which rely on the eigenstructure of a similarity matrix to partition points into disjoint clusters, with points in the same cluster having high...
A graphical model for predicting protein molecular function (2006)
Barbara E Engelhardt, Michael I Jordan, Steven E Brenner
We present a simple statistical model of molecular function evolution to predict protein function. The model description encodes general knowledge of how molecular function evolves within a...
Xuanlong Nguyen, Martin J. Wainwright, Michael I. Jordan
optimal quantization rules for some sequential decision problems
Learning spectral clustering, with application to speech separation (2006)
Francis R. Bach, Michael I. Jordan, Yoshua Bengio
Spectral clustering refers to a class of techniques which rely on the eigenstructure of a similarity matrix to partition points into disjoint clusters, with points in the same cluster having high...
Bayesian Multi-Population Haplotype Inference via a Hierarchical Dirichlet Process Mixture (2006)
Eric P. Xing, Kyung-Ah Sohn, Michael I. Jordan, Yee-Whye Teh
Uncovering the haplotypes of single nucleotide polymorphisms and their population demography is essential for many biological and medical applications. Methods for haplotype inference developed thus...
Variational Inference for Dirichlet Process Mixtures (2006)
David M. Blei, Michael I. Jordan
Dirichlet process (DP) mixture models are the cornerstone of nonparametric Bayesian statistics, and the development of Monte-Carlo Markov chain (MCMC) sampling methods for DP mixtures has enabled the...
Bayesian Multi-Population Haplotype Inference (2006)
Via Hierarchical Dirichlet, Eric P. Xing, Kyung-ah Sohn, Michael I. Jordan, Yee-whye Teh
Uncovering the haplotypes of single nucleotide polymorphisms and their population demography is essential for many biological and medical applications. Methods for haplotype inference developed thus...
A Graphical Model for Predicting Protein Molecular Function (2006)
Barbara Engelhardt Bee, Barbara E. Engelhardt, Michael I. Jordan, Steven E. Brenner
We present a simple statistical model of molecular function evolution to predict protein function. The model description encodes general knowledge of how molecular function evolves within a...
On Optimal Quantization Rules for Some Sequential Decision Problems (2006)
Xuanlong Nguyen, Martin J. Wainwright, Michael I. Jordan
We consider the problem of sequential decentralized detection, a problem that entails several interdependent choices: the choice of a stopping rule (specifying the sample size), a global decision...
Francis Bach Francis, Michael I. Jordan, Yoshua Bengio
Spectral clustering refers to a class of techniques which rely on the eigenstructure of a similarity matrix to partition points into disjoint clusters, with points in the same cluster having high...
Robust Design of Biological Experiments (2006)
Patrick Flaherty Eecs, Patrick Flaherty, Michael I. Jordan, Adam P. Arkin
We address the problem of robust, computationally-efficient design of biological experiments. Classical optimal experiment design methods have not been widely adopted in biological practice, in part...
Kernel Dimension Reduction in (2006)
Regression Kenji Fukumizu, Kenji Fukumizu, Francis R. Bach, Centre De Morphologie, Michael I. Jordan
We present a new methodology for su#cient dimension reduction (SDR). Our methodology derives directly from a formulation of SDR in terms of the conditional independence of the covariate X from the...
Structured prediction, dual extragradient and Bregman projections (2006)
Ben Taskar, Simon Lacoste-julien, Michael I. Jordan
We present a simple and scalable algorithm for maximum-margin estimation of structured output models, including an important class of Markov networks and combinatorial models. We formulate the...
Convex and Semi-Nonnegative Matrix Factorizations (2006)
Chris Ding, Tao Li, Michael I. Jordan
We present several new variations on the theme of nonnegative matrix factorization (NMF). Considering factorizations of the form X = F G T, we focus on algorithms in which G is restricted to contain...
In-network PCA and anomaly detection (2006)
Ling Huang, Michael I. Jordan, Anthony Joseph, Minos Garofalakis, Nina Taft
We consider the problem of network anomaly detection in large distributed systems. In this setting, Principal Component Analysis (PCA) has been proposed as a method for discovering anomalies by...
In-network PCA and anomaly detection (2006)
Ling Huang, Michael I. Jordan, Anthony Joseph, Minos Garofalakis, Nina Taft
We consider the problem of network anomaly detection in large distributed systems. In this setting, Principal Component Analysis (PCA) has been proposed as a method for discovering anomalies by...
Structured prediction, dual extragradient and Bregman projections (2006)
Ben Taskar, Simon Lacoste-julien, Michael I. Jordan, P. Bennett, Emilio Parrado-hernández
We present a simple and scalable algorithm for maximum-margin estimation of structured output models, including an important class of Markov networks and combinatorial models. We formulate the...
Xuanlong Nguyen, Martin J. Wainwright, Michael I. Jordan
optimal quantization rules for some sequential decision problems
In-network PCA and anomaly detection (2006)
Ling Huang, Michael I. Jordan, Anthony Joseph, Minos Garofalakis, Nina Taft
We consider the problem of network anomaly detection in large distributed systems. In this setting, Principal Component Analysis (PCA) has been proposed as a method for discovering anomalies by...
Advanced tools for operators at Amazon.com (2006)
Peter Bodík, O Fox, Michael I. Jordan, David Patterson, Ajit Banerjee, Ramesh Jagannathan, ...
Despite significant efforts in the field of Autonomic Computing, system operators will still play a critical role in administering Internet services for many years to come. However, very little is...
In-network PCA and anomaly detection (2006)
Michael Jordan, Anthony D. Joseph, All Rights Reserved, Ling Huang, Ling Huang, Michael I. Jordan, ...
Copyright © 2007, by the author(s).
Word Alignment via Quadratic Assignment (2006)
Lacoste-Julien, Simon, Taskar, Ben, Klein, Dan, Jordan, Michael I.
Recently, discriminative word alignment methods have achieved state-of-the-art accuracies by extending the range of information sources that can be easily incorporated into aligners. The chief...
Kernel dimension reduction in regression (2006)
Kenji Fukumizu, Francis R. Bach, Michael I. Jordan
Acknowledgements. The authors thank the editor and anonymous refer-ees for their helpful comments. The authors also thank Dr. Yoichi Nishiyama for his helpful comments on the uniform convergence of...
Structured prediction via the extragradient method (2006)
Ben Taskar, Simon Lacoste-julien, Michael I. Jordan
We present a simple and scalable algorithm for large-margin estimation of structured models, including an important class of Markov networks and combinatorial models. We formulate the estimation...
Structured Prediction via the Extragradient Method (2005)
Taskar, Ben, Lacoste-Julien, Simon, Jordan, Michael I.
We present a simple and scalable algorithm for large-margin estimation of structured models, including an important class of Markov networks and combinatorial models. We formulate the estimation...
On surrogate loss functions and $f$-divergences (2005)
Nguyen, XuanLong, Wainwright, Martin J., Jordan, Michael I.
The goal of binary classification is to estimate a discriminant function $\gamma$ from observations of covariate vectors and corresponding binary labels. We consider an elaboration of this problem in...
Protein Molecular Function Prediction by Bayesian Phylogenomics (2005)
Barbara E. Engelhardt, Michael I. Jordan, Kathryn E. Muratore, Steven E. Brenner
We present a statistical graphical model to infer specific molecular function for unannotated protein sequences using homology. Based on phylogenomic principles, SIFTER (Statistical Inference of...
Genome-Wide Requirements for Resistance to Functionally Distinct DNA-Damaging Agents (2005)
William Lee, Michael Proctor, Patrick Flaherty, Michael I. Jordan, Adam P. Arkin, ...
The mechanistic and therapeutic differences in the cellular response to DNA-damaging compounds are not completely understood, despite intense study. To expand our knowledge of DNA damage, we assayed...
The DLR Hierarchy of Approximate Inference (2005)
Rosen-Zvi, Michal, Jordan, Michael I., Yuille, Alan L
We propose a hierarchy for approximate inference based on the Dobrushin, Lanford, Ruelle (DLR) equations. This hierarchy includes existing algorithms, such as belief propagation, and also motivates...
Patrick Flaherty, Guri Giaever, Jochen Kumm, Michael I. Jordan, Adam P. Arkin
Motivation: In haploinsufficiency profiling data, pleiotropic genes are often misclassified by clustering algorithms that impose the constraint that a gene or experiment belong to only one cluster....
Sharing clusters among related groups: Hierarchical Dirichlet processes (2005)
Yee Whye Teh, Michael I. Jordan, Matthew J. Beal, David M. Blei
Bayesian model for clustering problems involving multiple groups of data. Each group of data is modeled with a mixture, with the number of components being open-ended and inferred automatically by...
A probabilistic interpretation of canonical correlation analysis (2005)
Francis R. Bach, Michael I. Jordan
We give a probabilistic interpretation of canonical correlation (CCA) analysis as a latent variable model for two Gaussian random vectors. Our interpretation is similar to the probabilistic...
Peter Bodík, Greg Friedman, Lukas Biewald, Helen Levine, George C, Kayur Patel, ...
Web applications suffer from software and configuration faults that lower their availability. Recovering from failure is dominated by the time interval between when these faults appear and when they...
Sharing clusters among related groups: Hierarchical Dirichlet processes (2005)
Yee Whye Teh, Michael I. Jordan, Matthew J. Beal, David M. Blei
Bayesian model for clustering problems involving multiple groups of data. Each group of data is modeled with a mixture, with the number of components being open-ended and inferred automatically by...
Variational inference for Dirichlet process mixtures (2005)
David M. Blei, Michael I. Jordan
Abstract. Dirichlet process (DP) mixture models are the cornerstone of nonparametric Bayesian statistics, and the development of Monte-Carlo Markov chain (MCMC) sampling methods for DP mixtures has...
Extensions of the informative vector machine (2005)
Neil D. Lawrence, John C. Platt, Michael I. Jordan
Abstract The informative vector machine (IVM) is a practical method for Gaussian process regression and classification. The IVM produces a sparse approximation to a Gaussian process by combining...
Extensions of the Informative Vector Machine (2005)
Neil D. Lawrence, John C. Platt, Michael I. Jordan
The informative vector machine (IVM) is a practical method for Gaussian process regression and classification. The IVM produces a sparse approximation to a Gaussian process by combining assumed...
On Divergences, Surrogate Loss Functions, and Decentralized Detection (2005)
Xuanlong Nguyen Computer, Xuanlong Nguyen, Martin J. Wainwright, Michael I. Jordan
We develop a general correspondence between a family of loss functions that act as surrogates to 0-1 loss, and the class of Ali-Silvey or f-divergence functionals. This correspondence provides the...
Computing Regularization Paths for Learning Multiple Kernels (2005)
Francis R. Bach, Romain Thibaux, Michael I. Jordan
The problem of learning a sparse conic combination of kernel functions or kernel matrices for classification or regression can be achieved via the regularization by a block 1-norm [1]. In this paper,...
Discriminative Training of Hidden Markov Models for Multiple Pitch Tracking (2005)
Francis R. Bach, Michael I. Jordan
We present a multiple pitch tracking algorithm that is based on direct probabilistic modeling of the spectrogram of the signal. The model is a factorial hidden Markov model whose parameters are...
Genome-Wide Requirements for Resistance (2005)
To Functionally Distinct, William Lee, Michael Proctor, Patrick Flaherty, Michael I. Jordan, ...
this paper are as follows: CSM2 (P40465), DDC1 (Q08949), ELG1 (Q12050), LTE1 (P07866), MAG1 (P22134), MEC3 (Q02574), MMS1 (Q06211), MMS4 (P38257), MPH1 (P40562), MRE11 (P32829), MUS81 (Q04149), PSO2...
Predictive Low-Rank Decomposition for Kernel Methods (2005)
Francis Bach Francis, Michael I. Jordan
Low-rank matrix decompositions are essential tools in the application of kernel methods to large-scale learning problems. These decompositions have generally been treated as black boxes---the...
Divergences, Surrogate Loss Functions and Experimental Design (2005)
Experimental Design, Xuanlong Nguyen, Martin J. Wainwright, Michael I. Jordan
In this paper, we provide a general theorem that establishes a correspondence between surrogate loss functions in classification and the family of f-divergences. Moreover, we provide constructive...
Xuanlong Nguyen, Martin J. Wainwright, Michael I. Jordan
We establish a general correspondence between two classes of statistical functions: AliSilvey distances (also known as f -divergences) and surrogate loss functions. Ali-Silvey distances play an...
Francis R. Bach, Michael I. Jordan
We give a probabilistic interpretation of canonical correlation (CCA) analysis as a latent variable model for two Gaussian random vectors. Our interpretation is similar to the probabilistic...
Combining Visualization and Statistical Analysis to Improve Operator (2005)
Confidence And Efficiency, Peter Bodík, Greg Friedman, Lukas Biewald, Helen Levine, George C, ...
Web applications suffer from software and configuration faults that lower their availability. Recovering from failure is dominated by the time interval between when these faults appear and when they...
Nonparametric Decentralized Detection Using Kernel Methods (2005)
Xuanlong Nguyen, Martin J. Wainwright, Michael I. Jordan
We consider the problem of decentralized detection under constraints on the number of bits that can be transmitted by each sensor. In contrast to most previous work, in which the joint distribution...
Protein Molecular Function Prediction (2005)
Bayesian Phylogenomics Barbara, Barbara E. Engelhardt, Michael I. Jordan, Kathryn E. Muratore, Steven E. Brenner
this report, we use only GO IDA- and IMP-derived annotations as observations for SIFTER, because of the high error rate and contradictions in the non-experimental annotations (i.e., all annotation...
Martin J. Wainwright, Michael I. Jordan
relaxation for approximate inference in discrete Markov random fields
Variational inference for Dirichlet process mixtures (2005)
David M. Blei, Michael I. Jordan
Dirichlet process (DP) mixture models are the cornerstone of nonparametric Bayesian statistics, and the development of Monte-Carlo Markov chain (MCMC) sampling methods for DP mixtures has enabled...
Protein molecular function prediction by Bayesian phylogenomics (2005)
Barbara E. Engelhardt, Michael I. Jordan, Kathryn E. Muratore, Steven E. Brenner
We present a statistical graphical model to infer specific molecular function for unannotated protein sequences using homology. Based on phylogenomic principles, SIFTER (Statistical Inference of...
A kernel-based learning approach to ad hoc sensor network localization (2005)
Xuanlong Nguyen, Michael I. Jordan, Bruno Sinopoli
We show that the coarse-grained and £ne-grained localization problems for ad hoc sensor networks can be posed and solved as a pattern recognition problem using kernel methods from statistical...
On divergences, surrogate loss functions and decentralized detection (2005)
Xuanlong Nguyen, Martin J. Wainwright, Michael I. Jordan
detection
Xuanlong Nguyen, Martin J. Wainwright, Michael I. Jordan
We establish a general correspondence between two classes of statistical functions: Ali-Silvey distances (also known as f-divergences) and surrogate loss functions. Ali-Silvey distances play an...
A latent variable model for chemogenomic profiling (2005)
Flaherty, Patrick, Giaever, Guri, Kumm, Jochen, Jordan, Michael I., Arkin, Adam P.
Motivation: In haploinsufficiency profiling data, pleiotropic genes are often misclassified by clustering algorithms that impose the constraint that a gene or experiment belong to only one cluster....
A latent variable model for chemogenomic profiling (2005)
Flaherty, Patrick, Giaever, Guri, Kumm, Jochen, Jordan, Michael I., Arkin, Adam P.
Motivation: In haploinsufficiency profiling (HIP) data (Giaever et al., 2002), pleiotropic genes are often misclassified by clustering algorithms that impose the constraint that a gene or experiment...
Subtree power analysis finds optimal species for comparative genomics (2004)
McAuliffe, Jon D., Jordan, Michael I., Pachter, Lior
Sequence comparison across multiple organisms aids in the detection of regions under selection. However, resource limitations require a prioritization of genomes to be sequenced. This prioritization...
Probabilistic Independence Networks for Hidden Markov Probability Models (2004)
Smyth, Padhraic, Heckerman, Cavid, Jordan, Michael I
In this paper we explore hidden Markov models(HMMs) and related structures within the general framework of probabilistic independence networks (PINs). The paper contains a self-contained review of...
A direct formulation for sparse PCA using semidefinite programming (2004)
D'Aspremont, Alexandre, Ghaoui, Laurent El, Jordan, Michael I., Lanckriet, Gert R. G.
We examine the problem of approximating, in the Frobenius-norm sense, a positive, semidefinite symmetric matrix by a rank-one matrix, with an upper bound on the cardinality of its eigenvector. The...
Statistical applications in fields such as bioinformatics, information retrieval, speech processing, image processing and communications often involve large-scale models in which thousands or...
Logos: A modular Bayesian model for de novo motif detection (2004)
Eric P. Xing, Michael I. Jordan
The complexity of the global organization and internal structures of motifs in higher eukaryotic organisms raises significant challenges for motif detection techniques. To achieve successful de novo...
Public deployment of cooperative bug isolation (2004)
Ben Liblit, Mayur Naik, Alice X. Zheng, Alex Aiken, Michael I. Jordan
As part of our work on Cooperative Bug Isolation (CBI) we have undertaken to instrument and distribute binaries for a number of large open source projects. This public deployment is an important step...
Hierarchical topic models and the nested Chinese restaurant process (2004)
David M. Blei, Thomas L. Griffiths, Michael I. Jordan, Joshua B. Tenenbaum
We address the problem of learning topic hierarchies from data. The model selection problem in this domain is daunting—which of the large collection of possible trees to use? We take a Bayesian...
Bayesian Haplotype Inference via the Dirichlet Process (2004)
Eric Xing, Roded Sharan, Michael I. Jordan
The problem of inferring haplotypes from genotypes of single nucleotide polymorphisms (SNPs) is essential for the understanding of genetic variation within and among populations, with important...
Learning in graphical models (2004)
Statistical applications in fields such as bioinformatics, information retrieval, speech processing, image processing and communications often involve large-scale models in which thousands or...
Decentralized detection and classification using kernel methods (2004)
Xuanlong Nguyen, Martin J. Wainwright, Michael I. Jordan
We consider the problem of decentralized detection under constraints on the number of bits that can be transmitted by each sensor. In contrast to most previous work, in which the joint distribution...
Dimensionality reduction for supervised learning with reproducing kernel Hilbert spaces (2004)
Kenji Fukumizu, Francis R. Bach, Michael I. Jordan, Chris Williams
We propose a novel method of dimensionality reduction for supervised learning problems. Given a regression or classification problem in which we wish to predict a response variable Y from an...
Hierarchical Dirichlet processes (2004)
Yee Whye Teh, Michael I. Jordan, Matthew J. Beal, David M. Blei
program. The authors wish to acknowledge helpful discussions with Lancelot James and Jim Pitman and the referees for useful comments. 1 We consider problems involving groups of data, where each...
Hierarchical Dirichlet processes (2004)
Yee Whye Teh, Michael I. Jordan, Matthew J. Beal, David M. Blei
ywteh,jordan,blei�
Bayesian Haplotype Inference via the Dirichlet Process (2004)
Eric P. Xing, Michael I. Jordan, Roded Sharan
The problem of inferring haplotypes from genotypes of single nucleotide polymorphisms (SNPs) is essential for the understanding of genetic variation within and among populations, with important...
Dimensionality reduction for supervised learning with reproducing kernel Hilbert spaces (2004)
Kenji Fukumizu, Francis R. Bach, Michael I. Jordan, Chris Williams
We propose a novel method of dimensionality reduction for supervised learning problems. Given a regression or classification problem in which we wish to predict a response variable Y from an...
Hierarchical Dirichlet processes (2004)
Yee Whye Teh, Michael I. Jordan, Matthew J. Beal, David M. Blei
1 We consider problems involving groups of data, where each observation within a group is a draw from a mixture model, and where it is desirable to share mixture components between groups. We assume...
Kalman filtering with intermittent observations (2004)
Bruno Sinopoli, Student Member, Luca Schenato, Massimo Franceschetti, Kameshwar Poolla, Michael I. Jordan, ...
Abstract—Motivated by navigation and tracking applications within sensor networks, we consider the problem of performing Kalman filtering with intermittent observations. When data travel along...
Hierarchical topic models and the nested Chinese restaurant process (2004)
David M. Blei, Thomas L. Griffiths, Michael I. Jordan, Joshua B. Tenenbaum
We address the problem of learning topic hierarchies from data. The model selection problem in this domain is daunting—which of the large collection of possible trees to use? We take a Bayesian...
Hierarchical Dirichlet processes (2004)
Yee Whye Teh, Michael I. Jordan, Matthew J. Beal, David M. Blei
1 We consider problems involving groups of data, where each observation within a group is a draw from a mixture model, and where it is desirable to share mixture components between groups. We assume...
Semiparametric Latent Factor Models (2004)
Matthias Seeger, Yee-whye Teh, Michael I. Jordan
We propose a semiparametric model for regression and classification problems involving multiple response variables. The model makes use of a set of Gaussian processes to model the relationship to the...
Jon D. Mcauliffe, Michael I. Jordan, Lior Pachter
Subtree power analysis finds optimal species for
Bayesian Haplotype Inference via the Dirichlet Process (2004)
Eric Xing, Roded Sharan, Michael I. Jordan
The problem of inferring haplotypes from genotypes of single nucleotide polymorphisms (SNPs) is essential for the understanding of genetic variation within and among populations, with important...
A direct formulation for sparse PCA using semidefinite programming (2004)
Laurent El Ghaoui, Michael I. Jordan
We examine the problem of approximating, in the Frobenius-norm sense, a positive, semidefinite symmetric matrix by a rank-one matrix, with an upper bound on the cardinality of its eigenvector. The...
Bayesian Haplotype Inference via the Dirichlet Process (2004)
Eric P. Xing, Roded Sharan, Michael I. Jordan
Abstract. The problem of inferring haplotypes from genotypes of single nucleotide polymorphisms (SNPs) is essential for the understanding of genetic variation within and among populations, with...
Fast kernel learning using sequential minimal optimizationTechnical Report CSD-04-1307 (2004)
Michael I, Francis R. Bach, Francis R. Bach, Michael I. Jordan
While classical kernel-based classifiers are based on a single kernel, in practice it is often desirable to base classifiers on combinations of multiple kernels. Lanckriet et al. (2004) considered...
Failure Diagnosis Using Decision Trees (2004)
Mike Chen Alice, Mike Chen, Alice X. Zheng, Jim Lloyd, Michael I. Jordan, Eric Brewer
We present a decision tree learning approach to diagnosing failures in large Internet sites. We record runtime properties of each request and apply automated machine learning and data mining...
LOGOS: A Modular Bayesian Model for de novo Motif Detection (2004)
Eric P. Xing, Wei Wu, Michael I. Jordan, Richard M. Karp
this paper, we present LOGOS,anintegratedLOcal and GlObal motif Sequence model for biopolymer sequences, which provides a principled framework for developing, modularizing, extending and computing...
Variational Methods for the Dirichlet Process (2004)
David M. Blei, Michael I. Jordan
Variational inference methods, including mean field methods and loopy belief propagation, have been widely used for approximate probabilistic inference in graphical models.
Vol No Pages, Tijl De Bie, Nello Cristianini, Michael I. Jordan, William Stafford Noble
Motivation: During the past decade, the new focus on genomics has highlighted a particular challenge: to integrate the different views of the genome that are provided by various types of experimental...
Large Margin Classifiers: Convex Loss, Low Noise, and Convergence Rates (2004)
Peter L. Bartlett, Michael I. Jordan, Jon M. Mcauliffe
Many classification algorithms, including the support vector machine, boosting and logistic regression, can be viewed as minimum contrast methods that minimize a convex surrogate of the 0-1 loss...
Nonparametric empirical Bayes for the Dirichlet process mixture model (2004)
Jon Mcauliffe David, David M. Blei, Michael I. Jordan
The Dirichlet process prior allows flexible nonparametric mixture modeling. The number of mixture components is not specified in advance and can grow as new data come in. However, the behavior of the...
Sparse Gaussian Process Classification with Multiple Classes (2004)
Matthias Seeger, Michael I. Jordan
Sparse approximations to Bayesian inference for nonparametric Gaussian Process models scale linearly in the number of training points, allowing for the application of these powerful kernel-based...
Large Margin Classifiers: Convex Loss, Low Noise, and Convergence Rates (2004)
Peter L. Bartlett, Michael I. Jordan, Jon D. Mcauliffe
Many classification algorithms, including the support vector machine, boosting and logistic regression, can be viewed as minimum contrast methods that minimize a convex surrogate of the 0-1 loss...
Decentralized Detection and Classification using Kernel Methods (2004)
Xuanlong Nguyen Xuanlong, Martin J. Wainwright, Michael I. Jordan
We consider the problem of decentralized detection under constraints on the number of bits that can be transmitted by each sensor. In contrast to most previous work, in which the joint distribution...
Gert Lanckriet Gert, Nello Cristianini, Peter Bartlett, Laurent El Ghaoui, Michael I. Jordan, Bernhard Schölkopf
Kernel-based learning algorithms work by embedding the data into a Euclidean space, and then searching for linear relations among the embedded data points. The embedding is performed implicitly, by...
Failure Diagnosis Using Decision Trees (2004)
Mike Chen Alice, Mike Chen, Alice X. Zheng, Jim Lloyd, Michael I. Jordan, Eric Brewer
We present a decision tree learning approach to diagnosing failures in large Internet sites. We record runtime properties of each request and apply automated machine learning and data mining...
Autonomous Helicopter Flight (2004)
Via Reinforcement Learning, Andrew Y. Ng, H. Jin Kim, Michael I. Jordan, Shankar Sastry
Autonomous helicopter flight represents a challenging control problem, with complex, noisy, dynamics. In this paper, we describe a successful application of reinforcement learning to autonomous...
Blind One-Microphone Speech Separation: A Spectral Learning Approach (2004)
Francis R. Bach, Michael I. Jordan
We present an algorithm to perform blind, one-microphone speech separation.
Treewidth-Based Conditions for Exactness (2004)
Martin J. Wainwright, Michael I. Jordan
The Sherali-Adams (SA) and Lasserre (LS) approaches are "lift-and-project" methods that generate nested sequences of linear and/or semidefinite relaxations of an arbitrary 0-1 polytope ....
Decentralized detection and classification using kernel methods (2004)
Xuanlong Nguyen, Martin J. Wainwright, Michael I. Jordan
We consider the problem of decentralized detection under constraints on the number of bits that can be transmitted by each sensor. In contrast to most previous work, in which the joint distribution...
Sparse Gaussian process classification with multiple classes (2004)
Matthias Seeger, Michael I. Jordan
Sparse approximations to Bayesian inference for nonparametric Gaussian Process models scale linearly in the number of training points, allowing for the application of these powerful kernel-based...
Inverted autonomous helicopter flight via reinforcement learning (2004)
Andrew Y. Ng, H. Jin Kim, Michael I. Jordan, Shankar Sastry
Autonomous helicopter flight represents a challenging control problem, with complex, noisy, dynamics. In this paper, we describe a successful application of reinforcement learning to autonomous...
Dimensionality reduction for supervised learning with reproducing kernel Hilbert spaces (2004)
Kenji Fukumizu, Francis R. Bach, Michael I. Jordan, Chris Williams
We propose a novel method of dimensionality reduction for supervised learning problems. Given a regression or classification problem in which we wish to predict a response variable Y from an...
Hierarchical Dirichlet processes (2004)
Yee Whye Teh, Michael I. Jordan, Matthew J. Beal, David M. Blei
1 We consider problems involving groups of data, where each observation within a group is a draw from a mixture model, and where it is desirable to share mixture components between groups. We assume...
Large margin classifiers: Convex loss, low noise and convergence rates (2004)
Peter L. Bartlett, Michael I. Jordan, Jon D. Mcauliffe
Many classification algorithms, including the support vector machine, boosting and logistic regression, can be viewed as minimum contrast methods that minimize a convex surrogate of the 0-1 loss...
Decentralized detection and classification using kernel methods (2004)
Martin J. Wainwright, Michael I. Jordan
We consider the problem of decentralized detection under constraints on the number of bits that can be transmitted by each sensor. In contrast to most previous work, in which the joint distribution...
Bayesian Haplotype Inference via the Dirichlet Process (2004)
Eric Xing, Roded Sharan, Michael I. Jordan
The problem of inferring haplotypes from genotypes of single nucleotide polymorphisms (SNPs) is essential for the understanding of genetic variation within and among populations, with important...
Kernel Dimensionality Reduction for Supervised Learning (2004)
Kenji Fukumizu, Francis R. Bach, Michael I. Jordan
We propose a novel method of dimensionality reduction for supervised learning. Given a regression or classification problem in which we wish to predict a variable Y from an explanatory vector X , we...
Bayesian Haplotype Inference via the Dirichlet Process (2004)
Eric P. Xing, Michael I. Jordan, Roded Sharan
The problem of inferring haplotypes from genotypes of single nucleotide polymorphisms (SNPs) is essential for the understanding of genetic variation within and among populations, with important...
Decentralized detection and classification using kernel methods (2004)
Xuanlong Nguyen, Martin J. Wainwright, Michael I. Jordan
We consider the problem of decentralized detection under constraints on the number of bits that can be transmitted by each sensor. In contrast to most previous work, in which the joint distribution...
Learning the kernel matrix with semidefinite programming (2004)
Nello Cristianini, Peter Bartlett, Laurent El Ghaoui, Michael I. Jordan, Bernhard Schölkopf
Kernel-based learning algorithms work by embedding the data into a Euclidean space, and then searching for linear relations among the embedded data points. The embedding is performed implicitly, by...
Learning the kernel matrix with semidefinite programming (2004)
Nello Cristianini, Peter Bartlett, Laurent El Ghaoui, Michael I. Jordan, Bernhard Schölkopf
Kernel-based learning algorithms work by embedding the data into a Euclidean space, and then searching for linear relations among the embedded data points. The embedding is performed implicitly, by...
A direct formulation for sparse PCA using semidefinite programming (2004)
Laurent El Ghaoui, Michael I. Jordan
Abstract. Given a covariance matrix, we consider the problem of maximizing the variance explained by a particular linear combination of the input variables while constraining the number of nonzero...
Kalman Filtering with Intermittent Observations (2004)
Bruno Sinopoli, Luca Schenato, Massimo Franceschetti, Kameshwar Poolla, Michael I. Jordan, Shankar S. Sastry
Motivated by navigation and tracking applications within sensor networks, we consider the problem of performing Kalman filtering with intermittent observations. When data travel along unreliable...
A direct formulation for sparse PCA using semidefinite programming (2004)
Laurent El Ghaoui, Michael I. Jordan
Abstract. Given a covariance matrix, we consider the problem of maximizing the variance explained by a particular linear combination of the input variables while constraining the number of nonzero...
Failure diagnosis using decision trees (2004)
Mike Chen, Alice X. Zheng, Jim Lloyd, Michael I. Jordan, Eric Brewer
We present a decision tree learning approach to diagnosing failures in large Internet sites. We record runtime properties of each request and apply automated machine learning and data mining...
Hierarchical Topic Models and the Nested Chinese Restaurant Process (2004)
David M. Blei, Thomas L. Griffiths, Michael I. Jordan, Joshua B. Tenenbaum
We address the problem of learning topic hierarchies from data. The model selection problem in this domain is daunting---which of the large collection of possible trees to use? We take a Bayesian...
A Kernel-Based Learning Approach to Ad Hoc Sensor Network Localization (2004)
Xuanlong Nguyen, Xuanlong Nguyen, Michael I. Jordan, Michael I. Jordan, Bruno Sinopoli, ...
We show that the coarse-grained and fine-grained localization problems for ad hoc sensor networks can be posed and solved as a pattern recognition problem using kernel methods from statistical...
Failure diagnosis using decision trees (2004)
Mike Chen, Alice X. Zheng, Jim Lloyd, Michael I. Jordan, Eric Brewer
We present a decision tree learning approach to diagnosing failures in large Internet sites. We record runtime properties of each request and apply automated machine learning and data mining...
Multiple-sequence functional annotation and the generalized hidden Markov phylogeny (2004)
McAuliffe, Jon D., Pachter, Lior, Jordan, Michael I.
Motivation: Phylogenetic shadowing is a comparative genomics principle which allows for the discovery of conserved regions in sequences from multiple closely-related organisms. We develop a formal...
A statistical framework for genomic data fusion (2004)
Lanckriet, Gert R. G., De Bie, Tijl, Cristianini, Nello, Jordan, Michael I., Noble, William Stafford
Motivation: During the past decade, the new focus on genomics has highlighted a particular challenge: to integrate the different views of the genome that are provided by various types of experimental...
A statistical framework for genomic data fusion (2004)
Lanckriet, Gert R. G., De Bie, Tijl, Cristianini, Nello, Jordan, Michael I., Noble, William Stafford
Motivation: During the past decade, the new focus on genomics has highlighted a particular challenge: to integrate the different views of the genome that are provided by various types of experimental...
Multiple-sequence functional annotation and the generalized hidden Markov phylogeny (2004)
McAuliffe, Jon D., Pachter, Lior, Jordan, Michael I.
Motivation: Phylogenetic shadowing is a comparative genomics principle that allows for the discovery of conserved regions in sequences from multiple closely related organisms. We develop a formal...
Multiple-sequence functional annotation and the generalized hidden Markov phylogeny (2004)
McAuliffe, Jon D., Pachter, Lior, Jordan, Michael I.
Motivation: Phylogenetic shadowing is a comparative genomics principle which allows for the discovery of conserved regions in sequences from multiple closely-related organisms. We develop a formal...
A statistical framework for genomic data fusion (2004)
Lanckriet, Gert R. G., De Bie, Tijl, Cristianini, Nello, Jordan, Michael I., Noble, William Stafford
Motivation: During the past decade, the new focus on genomics has highlighted a particular challenge: to integrate the different views of the genome that are provided by various types of experimental...
A direct formulation for sparse pca using semidefinite programming (2004)
Laurent El Ghaoui, Michael I. Jordan
Abstract. Given a covariance matrix, we consider the problem of maximizing the variance explained by a particular linear combination of the input variables while constraining the number of nonzero...
Analyse en composantes indépendantes et réseaux bayésiens (2003)
BACH, Francis R., JORDAN, Michael I.
- Une généralisation de l'analyse en composantes indépendantes (ACI) est introduite: au lieu de déterminer une application linéaire qui rend les composantes indépendantes, nous cherchons une...
Graph partition strategies for generalized mean field inference (2003)
Eric P. Xing, Michael I. Jordan, Stuart Russell
An autonomous variational inference algorithm for arbitrary graphical models requires the ability to optimize variational approximations over the space of model parameters as well as over the choice...
Beyond independent components: trees and clusters (2003)
Francis R. Bach, Michael I. Jordan
We present a generalization of independent component analysis (ICA), where instead of looking for a linear transform that makes the data components independent, we look for a transform that makes the...
Latent dirichlet allocation (2003)
David M. Blei, Andrew Y. Ng, Michael I. Jordan, John Lafferty
We describe latent Dirichlet allocation (LDA), a generative probabilistic model for collections of discrete data such as text corpora. LDA is a three-level hierarchical Bayesian model, in which each...
Matching words and pictures (2003)
Kobus Barnard, Pinar Duygulu, David Forsyth, Nando De Freitas, David M. Blei, Michael I. Jordan, ...
We present a new approach for modeling multi-modal data sets, focusing on the specific case of segmented images with associated text. Learning the joint distribution of image regions and words has...
Matching words and pictures (2003)
Kobus Barnard, Pinar Duygulu, David Forsyth, Nando De Freitas, David M. Blei, Michael I. Jordan, ...
We present a new approach for modeling multi-modal data sets, focusing on the specific case of segmented images with associated text. Learning the joint distribution of image regions and words has...
Variational inference in a truncated Dirichlet process (2003)
David M. Blei, Michael I. Jordan
The N-component truncated Dirichlet process (DPN) is defined in Ishwaran and James [2001] and converges almost surely to a true Dirichlet process (DP∞). Like a full Dirichlet process, this...
Matching words and pictures (2003)
Kobus Barnard, Pinar Duygulu, David Forsyth, Nando De Freitas, David M. Blei, Michael I. Jordan, ...
We present a new approach for modeling multi-modal data sets, focusing on the specific case of segmented images with associated text. Learning the joint distribution of image regions and words has...
A hierarchical Bayesian Markovian model for motifs in biopolymer sequences (2003)
Eric P. Xing, Michael I. Jordan, Richard M. Karp, Stuart Russell
We propose a dynamic Bayesian model for motifs in biopolymer sequences which captures rich biological prior knowledge and positional dependencies in motif structure in a principled way. Our model...
Bug isolation via remote program sampling (2003)
Ben Liblit, Alex Aiken, Alice X. Zheng, Michael I. Jordan
We propose a low-overhead sampling infrastructure for gathering information from the executions experienced by a program 's user community. Several example applications illustrate ways to use...
Convexity, classification, and risk bounds (2003)
Peter L. Bartlett, Michael I. Jordan, Jon D. Mcauliffe
Many of the classification algorithms developed in the machine learning literature, including the support vector machine and boosting, can be viewed as minimum contrast methods that minimize a convex...
Latent dirichlet allocation (2003)
David M. Blei, Andrew Y. Ng, Michael I. Jordan
We propose a generative model for text and other collections of discrete data that generalizes or improves on several previous models including naive Bayes/unigram, mixture of unigrams [6], and...
Distance metric learning, with application to clustering with side-information (2003)
Eric P. Xing, Andrew Y. Ng, Michael I. Jordan, Stuart Russell
Many algorithms rely critically on being given a good metric over their inputs. For instance, data can often be clustered in many "plausible" ways, and if a clustering algorithm...
Latent dirichlet allocation (2003)
David M. Blei, Andrew Y. Ng, Michael I. Jordan
We propose a generative model for text and other collections of discrete data, that generalizes or improves on several previous models, including naive Bayes/unigram, mixtures of naive Bayes [6], and...
Matching words and pictures (2003)
Kobus Barnard, Pinar Duygulu, David Forsyth, Nando De Freitas, David M. Blei, Michael I. Jordan, ...
We present a new approach for modeling multi-modal data sets, focusing on the specific case of segmented images with associated text. Learning the joint distribution of image regions and words has...
Statistical Debugging of Sampled Programs (2003)
Alice Zheng Ee, Alice X. Zheng, Ben Liblit, Michael I. Jordan, Alex Aiken
We present a novel strategy for automatically debugging programs given sampled data from thousands of actual user runs. Our goal is to pinpoint those features that are most correlated with crashes....
On the Concentration of Expectation and Approximate Inference in Layered Networks (2003)
Xuanlong Nguyen, Michael I. Jordan
We present an analysis of concentration-of-expectation phenomena in layered Bayesian networks that use generalized linear models as the local conditional probabilities. This framework encompasses a...
Latent Dirichlet Allocation (2003)
David M. Blei, Andrew Y. Ng, Michael I. Jordan
We describe latent Dirichlet allocation (LDA), a generative probabilistic model for collections of discrete data such as text corpora. LDA is a three-level hierarchical Bayesian model, in which each...
A Hierarchical Bayesian Markovian Model for Motifs in Biopolymer Sequences (2003)
Eric P. Xing, Michael I. Jordan, Richard M. Karp, Stuart Russell
We propose a dynamic Bayesian model for motifs in biopolymer sequences which captures rich biological prior knowledge and positional dependencies in motif structure in a principled way. Our model...
On the Concentration of Expectation and Approximate Inference in Layered Networks (2003)
XuanLong Nguyen, Michael I. Jordan
We present an analysis of concentration-of-expectation phenomena in layered Bayesian networks that use generalized linear models as the local conditional probabilities. This framework encompasses a...
Hierarchical Dirichlet Processes (2003)
Yee Whye Teh, Matthew J. Beal, Michael I. Jordan, David M. Blei
We consider problems involving groups of data, where each observation within a group is a draw from a mixture model, and where it is desirable to share mixture components both within and between...
Variational inference in graphical models: The view from the marginal polytope (2003)
Martin J. Wainwright, Michael I. Jordan
Underlying a variety of techniques for approximate inference---among them mean field, sum-product, and cluster variational methods---is a classical variational principle from statistical physics,...
Finding Clusters In Independent Component Analysis (2003)
Francis R. Bach, Michael I. Jordan
We present a class of algorithms that find clusters in independent component analysis: the data are linearly transformed so that the resulting components can be grouped into clusters, such that...
Learning Graphical Models (2003)
With Mercer Kernels, Francis R. Bach, Michael I. Jordan
We present a class of algorithms for learning the structure of graphical models from data. The algorithms are based on a measure known as the kernel generalized variance (KGV), which essentially...
LOGOS: a modular Bayesian model for de novo motif detection (2003)
Eric P. Xing, Michael I. Jordan, Wei Wu
The complexity of the global organization and internal structures of motifs in higher eukaryotic organisms raises significant challenges for motif detection techniques. To achieve successful de novo...
Bug Isolation via Remote Program Sampling (2003)
Ben Liblit Alex, Alice X. Zheng, Alex Aiken, Michael I. Jordan
We propose a low-overhead sampling infrastructure for gathering information from the executions experienced by a program 's user community. Several example applications illustrate ways to use...
Hierarchical Dirichlet Processes (2003)
Yee Whye Teh, Michael I. Jordan, Matthew J. Beal, David M. Blei
We consider problems involving groups of data, where each observation within a group is a draw from a mixture model, and where it is desirable to share mixture components between groups. We assume...
Kalman Filtering with Intermittent Observations* (2003)
Bruno Sinopoli Luca, Luca Schenato, Massimo Franceschetti, Kameshwar Poolla, Michael I. Jordan, Shankar S. Sastry
Motivated by our experience in building sensor networks for navigation as part of the Networked Embedded Systems Technology (NEST) project at Berkeley, we consider the problem of performing Kalman...
Martin Wainwright And, Martin J. Wainwright, Martin J. Wainwright, Michael I. Jordan, Michael I. Jordan
We present a new method for calculating approximate marginals for probability distributions defined by graphs with cycles, based on a Gaussian entropy bound combined with a semidefinite outer bound...
Hierarchical Dirichlet Processes (2003)
Yee Whye Teh, Michael I. Jordan, Matthew J. Beal, David M. Blei
We propose the hierarchical Dirichlet process (HDP), a nonparametric Bayesian model for clustering problems involving multiple groups of data. Each group of data is modeled with a mixture, with the...
Michael Jordan Computer, Michael I. Jordan
Statistical applications in fields such as bioinformatics, information retrieval, speech processing, image processing and communications often involve large-scale models in which thousands or...
Hierarchical Dirichlet Processes (2003)
Yee Whye Teh, Michael I. Jordan, Matthew J. Beal, David M. Blei
We consider problems involving groups of data, where each observation within a group is a draw from a mixture model, and where it is desirable to share mixture components between groups. We assume...
A hierarchical Bayesian Markovian model for motifs in biopolymer sequences (2003)
Eric P. Xing, Michael I. Jordan, Richard M. Karp, Stuart Russell
We propose a dynamic Bayesian model for motifs in biopolymer sequences which captures rich biological prior knowledge and positional dependencies in motif structure in a principled way. Our model...
A hierarchical Bayesian Markovian model for motifs in biopolymer sequences (2003)
Eric P. Xing, Michael I. Jordan, Richard M. Karp, Stuart Russell
We propose a dynamic Bayesian model for motifs in biopolymer sequences which captures rich biological prior knowledge and positional dependencies in motif structure in a principled way. Our model...
Distance metric learning, with application to clustering with side-information (2003)
Eric P. Xing, Andrew Y. Ng, Michael I. Jordan, Stuart Russell
Many algorithms rely critically on being given a good metric over their inputs. For instance, data can often be clustered in many “plausible” ways, and if a clustering algorithm such as K-means...
Convexity, classification, and risk bounds (2003)
Peter L. Bartlett, Michael I. Jordan, Jon D. Mcauliffe
Many of the classification algorithms developed in the machine learning literature, including the support vector machine and boosting, can be viewed as minimum contrast methods that minimize a convex...
A hierarchical Bayesian Markovian model for motifs in biopolymer sequences (2003)
Eric P. Xing, Michael I. Jordan, Richard M. Karp, Stuart Russellcomputer, Science Division
Abstract We propose a dynamic Bayesian model for motifs in biopolymer se-quences which captures rich biological prior knowledge and positional dependencies in motif structure in a principled way. Our...
Distance metric learning, with application to clustering with side-information (2003)
Eric P. Xing, Andrew Y. Ng, Michael I. Jordan, Stuart Russell
Many algorithms rely critically on being given a good metric over their inputs. For instance, data can often be clustered in many “plausible” ways, and if a clustering algorithm such as K-means...
Distance Metric Learning, with Application to Clustering with Side-information (2003)
Eric P. Xing, Andrew Y. Ng, Michael I. Jordan, Stuart Russell
Many algorithms rely critically on being given a good metric over their inputs. For instance, data can often be clustered in many "plausible" ways, and if a clustering algorithm such as...
A Generalized Mean Field Algorithm for Variational Inference in Exponential Families (2003)
Eric P. Xing, Michael I. Jordan, Stuart Russell
We present a class of generalized mean field (GMF) algorithms for approximate inference in exponential family graphical models which is analogous to the generalized belief propagation (GBP) or...
Robust Novelty Detection with Single-Class MPM (2003)
Laurent El Ghaoui, Michael I. Jordan
In this paper we consider the problem of novelty detection, presenting an algorithm that aims to nd a minimal region in input space containing a fraction of the probability mass underlying a data...
A hierarchical Bayesian Markovian model for motifs in biopolymer sequences (2003)
Eric P. Xing, Michael I. Jordan, Richard M. Karp, Stuart Russell
We propose a dynamic Bayesian model for motifs in biopolymer sequences which captures rich biological prior knowledge and positional dependencies in motif structure in a principled way. Our model...
Distance metric learning, with application to clustering with side-information (2003)
Eric P. Xing, Andrew Y. Ng, Michael I. Jordan, Stuart Russell
Many algorithms rely critically on being given a good metric over their inputs. For instance, data can often be clustered in many “plausible” ways, and if a clustering algorithm such as K-means...
Kenji Fukumizu, Francis R. Bach, Michael I. Jordan
We propose a novel method of dimensionality reduction for supervised learning problems. Given a regression or classification problem in which we wish to predict a response variable Y from an...
Convexity, classification, and risk bounds (2003)
Peter L. Bartlett, Michael I. Jordan, Jon D. Mcauliffe
Many of the classification algorithms developed in the machine learning literature, including the support vector machine and boosting, can be viewed as minimum contrast methods that minimize a convex...
Latent dirichlet allocation (2003)
David M. Blei, Andrew Y. Ng, Michael I. Jordan, John Lafferty
We describe latent Dirichlet allocation (LDA), a generative probabilistic model for collections of discrete data such as text corpora. LDA is a three-level hierarchical Bayesian model, in which each...
Sampling User Executions for Bug Isolation (2003)
Ben Liblit Alex, Alex Aiken, Alice X. Zheng, Michael I. Jordan
Introduction Many computer scientists think of a program as either correct (i.e. it meets some specification) or incorrect (i.e. it does not meet some specification). But industrial software...
Hierarchical Bayesian Models for Applications in Information retrieval (2003)
J. M. Bernardo, M. J. Bayarri, J. O. Berger, A. P. Dawid, D. Heckerman, ...
this article, we find that most of the factors are very close to # while four of the factors achieve significant expected counts. Looking at the distribution over words, z), for those four factors,...
Modeling Annotated Data (2003)
David M. Blei, Michael I. Jordan
We consider the problem of modeling annotated data---data with multiple types where the instance of one type (such as a caption) serves as a description of the other type (such as an image). We...
LOGOS: A Modular Bayesian Model for de novo Motif Detection (2003)
Eric P. Xing, Wei Wu, Michael I. Jordan, Richard M. Karp
The complexity of the global organization and internal structures of motifs in higher eukaryotic organisms raises significant challenges for motif detection techniques. To achieve successful de novo...
Latent Dirichlet Allocation (2003)
David M. Blei, Andrew Y. Ng, Michael I. Jordan, John Lafferty
We describe latent Dirichlet allocation (LDA), a generative probabilistic model for collections of discrete data such as text corpora. LDA is a three-level hierarchical Bayesian model, in which each...
Latent dirichlet allocation (2003)
David M. Blei, Andrew Y. Ng, Michael I. Jordan, John Lafferty
We describe latent Dirichlet allocation (LDA), a generative probabilistic model for collections of discrete data such as text corpora. LDA is a three-level hierarchical Bayesian model, in which each...
Learning Spectral Clustering (2003)
Michael I. Jordan, Francis R. Bach, Francis R. Bach
Spectral clustering refers to a class of techniques which rely on the eigenstructure of a similarity matrix to partition points into disjoint clusters, with points in the same cluster having high...
A Hierarchical Bayesian Markovian Model for Motifs in Biopolymer Sequences (2003)
Eric P. Xing, Michael I. Jordan, Richard M. Karp, Stuart Russell
We propose a dynamic Bayesian model for motifs in biopolymer sequences which captures rich biological prior knowledge and positional dependencies in motif structure in a principled way. Our model...
Bug Isolation via Remote Program Sampling (2003)
Ben Liblit Alex, Alice X. Zheng, Alex Aiken, Michael I. Jordan
We propose a low-overhead sampling infrastructure for gathering information from the executions experienced by a program 's user community. Several example applications illustrate ways to use...
Learning Spectral Clustering (2003)
Francis R. Bach, Michael I. Jordan
Spectral clustering refers to a class of techniques which rely on the eigenstructure of a similarity matrix to partition points into disjoint clusters with points in the same cluster having high...
Kernel Dimensionality Reduction for Supervised (2003)
Learning Kenji Fukumizu, Kenji Fukumizu, Francis R. Bach, Michael I. Jordan
We propose a novel method of dimensionality reduction for supervised learning. Given a regression or classification problem in which we wish to predict a variable Y from an explanatory vector X , we...
Hierarchical Topic Models and the Nested Chinese Restaurant Process (2003)
David M. Blei, Thomas L. Griffiths, Michael I. Jordan, Joshua B. Tenenbaum
We address the problem of learning topic hierarchies from data. The model selection problem in this domain is daunting -- which of the large collection of possible trees to use? We take a Bayesian...
Statistical Debugging of Sampled Programs (2003)
Alice Zheng Ee, Alice X. Zheng, Michael I. Jordan, Ben Liblit, Alex Aiken
We present a novel strategy for automatically debugging programs given sampled data from thousands of actual user runs. Our goal is to pinpoint those features that are most correlated with crashes....
Autonomous Helicopter Flight (2003)
Via Reinforcement Learning, Andrew Y. Ng, H. Jin Kim, Michael I. Jordan, Shankar Sastry
Autonomous helicopter flight represents a challenging control problem, with complex, noisy, dynamics. In this paper, we describe a successful application of reinforcement learning to autonomous...
On the Concentration of Expectation and Approximate Inference in Layered Networks (2003)
XuanLong Nguyen, Michael I. Jordan
We present an analysis of concentration-of-expectation phenomena in layered Bayesian networks that use generalized linear models as the local conditional probabilities. This framework encompasses a...
Semidefinite Relaxations for Approximate (2003)
Inference On Graphs, Martin J. Wainwright, Michael I. Jordan
We present a new method for calculating approximate marginals for probability distributions defined by graphs with cycles, based on a Gaussian entropy bound combined with a semidefinite outer bound...
Graph Partition Strategies for Generalized Mean Field Inference (2003)
Eric P. Xing, Eric P. Xing, Stuart Russell, Michael I. Jordan, Michael I. Jordan
An autonomous variational inference algorithm for arbitrary graphical model requires the ability to optimize variational approximations over the space of model parameters as well as over the choice...
Beyond Independent Components: Trees and Clusters (2003)
Francis R. Bach, Michael I. Jordan, Te-won Lee, Jean-francois Cardoso, Erkki Oja, Shun-ichi Amari
We present a generalization of independent component analysis (ICA), where instead of looking for a linear transform that makes the data components independent, we look for a transform that makes the...
Modeling Annotated Data (2003)
David M. Blei, Michael I. Jordan
We consider the problem of modeling annotated data---data with multiple types where the instance of one type (such as a caption) serves as a description of the other type (such as an image). We...
Semidefinite relaxations for approximate inference on graphs with cycles (2003)
Martin J. Wainwright, Martin J. Wainwright, Michael I. Jordan, Michael I. Jordan
graphs with cycles
Percy Liang, Dan Klein, Michael I. Jordan
The learning of probabilistic models with many hidden variables and nondecomposable dependencies is an important and challenging problem. In contrast to traditional approaches based on approximate...
The Timing of Endpoints in Movement (2002)
The issue of temporal control of motor behavior was investigated using a rhythmic tapping task. It was found that: (1) subjects are better able to tap before a beat than after a beat; (2) the...
Random Sampling of a Continuous-time Stochastic Dynamical System (2002)
Mario Micheli, Michael I. Jordan
We consider a dynamical system where the state equation is given by a linear stochastic differential equation and noisy measurements occur at discrete times, in correspondence of the arrivals of a...
Minimax probability machine (2002)
Laurent El Ghaoui, Chiranjib Bhattacharyya, Michael I. Jordan
When constructing a classier, the probability of correct classi-cation of future data points should be maximized. In the current paper this desideratum is translated in a very direct way into an...
Francis R. Bach, Michael I. Jordan
We present an algorithm that induces a class of models with thin junction trees---models that are characterized by an upper bound on the size of the maximal cliques of their triangulated graph. By...
Francis R. Bach, Michael I. Jordan
We present an algorithm that induces a class of models with thin junction trees---models that are characterized by an upper bound on the size of the maximal cliques of their triangulated graph. By...
Minimax probability machine (2002)
Laurent El Ghaoui, Chiranjib Bhattacharyya, Michael I. Jordan
When constructing a classifier, the probability of correct classification of future data points should be maximized. In the current paper this desideratum is translated in a very direct way into an...
Kernel independent component analysis (2002)
Michael I. Jordan, Francis R. Bach, Francis R. Bach
We present a class of algorithms for Independent Component Analysis (ICA) which use contrast functions based on canonical correlations in a reproducing kernel Hilbert space. On the one hand, we show...
Random Sampling of a Continuous-time (2002)
Stochastic Dynamical System, Mario Micheli, Michael I. Jordan
We consider a dynamical system where the state equation is given by a linear stochastic di#erential equation and noisy measurements occur at discrete times, in correspondence of the arrivals of a...
A Robust Minimax Approach to Classification (2002)
Laurent E Ghaoui, Chiranjib Bhattacharyya, Michael I. Jordan
When constructing a classifier, the probability of correct classification of future data points should be maximized. We consider a binary classification problem where the mean and covariance matrix...
Loopy Belief Propagation and Gibbs Measures (2002)
Sekhar C. Tatikonda, Michael I. Jordan
We address the question of convergence in the loopy belief propagation (LBP) algorithm.
Distance Metric Learning, with Application to Clustering with Side-information (2002)
Eric P. Xing, Andrew Y. Ng, Michael I. Jordan, Stuart Russell
Many algorithms rely critically on being given a good metric over their inputs. For instance, data can often be clustered in many "plausible" ways, and if a clustering algorithm such as...
Kernel independent component analysis (2002)
Francis R. Bach, Michael I. Jordan
We present a class of algorithms for independent component analysis (ICA) which use contrast functions based on canonical correlations in a reproducing kernel Hilbert space. On the one hand, we show...
A Minimal Intervention Principle for Coordinated Movement (2002)
Emanuel Todorov, Michael I. Jordan
Behavioral goals are achieved reliably and repeatedly with movements rarely reproducible in their detail. Here we offer an explanation: we show that not only are variability and goal achievement...
Random Sampling of a Continuous-time Stochastic Dynamical System (2002)
Mario Micheli, Michael I. Jordan
We consider a dynamical system where the state equation is given by a linear stochastic differential equation and noisy measurements occur at discrete times, in correspondence of the arrivals of a...
A robust minimax approach to classification (2002)
Laurent El Ghaoui, Chiranjib Bhattacharyya, Michael I. Jordan, Bernhard Schölkopf
When constructing a classifier, the probability of correct classification of future data points should be maximized. We consider a binary classification problem where the mean and covariance matrix...
A robust minimax approach to classification (2002)
Laurent El Ghaoui, Chiranjib Bhattacharyya, Michael I. Jordan, Bernhard Schölkopf
When constructing a classifier, the probability of correct classification of future data points should be maximized. We consider a binary classification problem where the mean and covariance matrix...
Tree-dependent Component Analysis (2002)
Francis R. Bach, Michael I. Jordan
We present a generalization of independent component analysis (ICA), where instead of looking for a linear transform that makes the data components independent, we look for a transform that makes the...
Kernel Independent Component Analysis (2002)
Francis R. Bach, Michael I. Jordan
We present a class of algorithms for independent component analysis (ICA) which use contrast functions based on canonical correlations in a reproducing kernel Hilbert space.
On spectral clustering: Analysis and an algorithm (2001)
Andrew Y. Ng, Michael I. Jordan, Yair Weiss
Despite many empirical successes of spectral clustering methods| algorithms that cluster points using eigenvectors of matrices derived from the data|there are several unresolved issues. First, there...
analysis, eigenvectors and stability (2001)
Andrew Y. Ng, Alice X. Zheng, Michael I. Jordan
The HITS and the PageRank algorithms are eigenvector methods for identifying "authoritative " or "influential " articles, given hyperlink or citation information....
Feature selection for high-dimensional genomic microarray data (2001)
Eric P. Xing, Michael I. Jordan, Richard M. Karp
We report on the successful application of feature selection methods to a classification problem in molecular biology involving only 72 data points in a 7130 dimensional space. Our approach is a...
Feature selection for high-dimensional genomic microarray data (2001)
Eric P. Xing, Michael I. Jordan, Richard M. Karp
We report on the successful application of feature selection methods to a classication problem in molecular biology involving only 72 data points in a 7130 dimensional space. Our approach is a hybrid...
Stable algorithms for link analysis (2001)
Andrew Y. Ng, Alice X. Zheng, Michael I. Jordan
The Kleinberg HITS and the Google PageRank algorithms are eigenvector methods for identifying "authoritative " or "influential " articles, given hyperlink or...
Andrew Y. Ng, Michael I. Jordan
The Gibbs classifier is a simple approximation to the Bayesian optimal classifier in which one samples from the posterior for the parameter `, and then classifies using the single classifier indexed...
analysis, eigenvectors and stability (2001)
Andrew Y. Ng, Alice X. Zheng, Michael I. Jordan
The HITS and the PageRank algorithms are eigenvector methods for identifying "authoritative " or "influential " articles, given hyperlink or citation information....
Stable algorithms for link analysis (2001)
Andrew Y. Ng, Alice X. Zheng, Michael I. Jordan
The Kleinberg HITS and the Google PageRank algorithms are eigenvector methods for identifying "authoritative " or "influential " articles, given hyperlink or...
Feature selection for high-dimensional genomic microarray data (2001)
Eric P. Xing, Michael I. Jordan, Richard M. Karp
We report on the successful application of feature selection methods to a classification problem in molecular biology involving only 72 data points in a 7130 dimensional space. Our approach is a...
Kernel Independent Component Analysis (2001)
Francis R. Bach, Michael I. Jordan
We present a class of algorithms for independent component analysis (ICA) which use contrast functions based on canonical correlations in a reproducing kernel Hilbert space. On the one hand, we show...
Feature selection for high-dimensional genomic microarray data (2001)
Eric P. Xing, Michael I. Jordan, Richard M. Karp
We report on the successful application of feature selection methods to a classification problem in molecular biology involving only 72 data points in a 7130 dimensional space. Our approach is a...
Learning with mixtures of trees (2000)
Marina Meilă, Michael I. Jordan
This paper describes the mixtures-of-trees model, a probabilistic model for discrete multidimensional domains. Mixtures-of-trees generalize the probabilistic trees of Chow and Liu [6] in a different...
Learning with mixtures of trees (2000)
Marina Meila, Michael I. Jordan
This paper describes the mixtures-of-trees model, a probabilistic model for discrete multidimensional domains. Mixtures-of-trees generalize the probabilistic trees of Chow and Liu (1968) in a...
Learning with mixtures of trees (2000)
Marina Meila, Michael I. Jordan
This paper describes the mixtures-of-trees model, a probabilistic model for discrete multidimensional domains. Mixtures-of-trees generalize the probabilistic trees of Chow and Liu [6] in a di#erent...
Bayesian parameter estimation via variational methods (2000)
Tommi S. Jaakkola, Michael I. Jordan
We consider a logistic regression model with a Gaussian prior distribution over the parameters. We show that an accurate variational transformation can be used to obtain a closed form approximation...
Approximate inference algorithms for two-layer Bayesian networks (2000)
Andrew Ng Computer, Andrew Y. Ng, Michael I. Jordan
We present a class of approximate inference algorithms for graphical models of the QMR-DT type. We give convergence rates for these algorithms and for the Jaakkola and Jordan (1999) algorithm, and...
Learning with mixtures of trees (2000)
Marina Meilă, Michael I. Jordan, Pack Kaelbling
This paper describes the mixtures-of-trees model, a probabilistic model for discrete multidimensional domains. Mixtures-of-trees generalize the probabilistic trees of Chow and Liu (1968) in a...
Learning with mixtures of trees (2000)
Marina Meilă, Michael I. Jordan, Pack Kaelbling
This paper describes the mixtures-of-trees model, a probabilistic model for discrete multidimensional domains. Mixtures-of-trees generalize the probabilistic trees of Chow and Liu (1968) in a...
Learning with mixtures of trees (2000)
Marina Meilă, Michael I. Jordan, Pack Kaelbling
This paper describes the mixtures-of-trees model, a probabilistic model for discrete multidimensional domains. Mixtures-of-trees generalize the probabilistic trees of Chow and Liu (1968) in a...
Michael I. Jordan, Arthur C. Smith
Learning with Mixtures of Trees by
Variational methods and the QMR-DT database (1999)
Tommi S. Jaakkola, Michael I. Jordan
We describe variational approximation methods for e cient probabilistic reasoning, applying these methods to the problem of diagnostic inference in the QMR-DT database. The QMR-DT database is a...
Variational methods and the QMR-DT database (1999)
Tommi S. Jaakkola, Michael I. Jordan
We describe variational approximation methods for efficient probabilistic reasoning, applying these methods to the problem of diagnostic inference in the QMR-DT database. The QMR-DT database is a...
Variational Probabilistic Inference and the QMR-DT Network (1999)
Tommi S. Jaakkola, Michael I. Jordan
We describe a variational approximation method for efficient inference in large-scale probabilistic models. Variational methods are deterministic procedures that provide approximations to marginal...
Loopy belief propagation for approximate inference: An empirical study (1999)
Kevin P. Murphy, Yair Weiss, Michael I. Jordan
Recently, researchers have demonstrated that "loopy belief propagation "--- the use of Pearl's polytree algorithm in a Bayesian network with loops--- can perform well in the...
Variational Probabilistic Inference and the QMR-DT Network (1999)
Tommi S. Jaakkola, Michael I. Jordan
We describe a variational approximation method for efficient inference in large-scale probabilistic models. Variational methods are deterministic procedures that provide approximations to marginal...
Loopy Belief Propagation for Approximate Inference: An Empirical Study (1999)
Kevin P. Murphy, Yair Weiss, Michael I. Jordan
Recently, a number of researchers have demonstrated excellent performance by using "loopy belief propagation" --- using Pearl's polytree algorithm in a Bayesian network with loops. The...
Variational probabilistic inference and the QMR-DT database (1999)
Tommi Jaakkola, Michael I. Jordan
We describe a variational approximation method for efficient inference in large-scale probabilistic models. Variational methods are deterministic procedures that provide approximations to marginal...
An introduction to variational methods for graphical models (1999)
This paper presents a tutorial introduction to the use of variational methods for inference and learning in graphical models (Bayesian networks and Markov random fields). We present a number of...
Variational Probabilistic Inference and the QMR-DT Network (1999)
Tommi S. Jaakkola, Michael I. Jordan
We describe a variational approximation method for e cient inference in large-scale probabilistic models. Variational methods are deterministic procedures that provide approximations to marginal and...
Lawrence K. Saul, Michael I. Jordan, Padhraic Smyth
Abstract. We study Markov models whose state spaces arise from the Cartesian product of two or more discrete random variables. We show how to parameterize the transition matrices of these models as a...
Hierarchical Mixtures of Experts and the EM Algorithm. (1998)
Jordan, Michael I., Jacobs, Robert A.
We present a tree-structured architecture for supervised learning. ne statistical model underlying the architecture is a hierarchical mixture model in which both the mixture coefficients and the...
Active Learning with Statistical Models. (1998)
Cohn, David A., Ghahramani, Zoubin, Jordan, Michael I.
For many types of learners one can compute the statistically optimal' way to select data. We review how these techniques have been used with feedforward neural networks. We then show how the same...
Learning from Incomplete Data. (1998)
Ghahramani, Zoubin, Jordan, Michael I.
Real-world learning tasks often involve high-dimensional data sets with complex patterns of missing features. In this paper we review the problem of learning from incomplete data from two statistical...
Fast Learning by Bounding Likelihoods in Sigmoid Type Belief Networks. (1998)
Jaakkola, Tommi S., Saul, Lawrence K., Jordan, Michael I.
Sigmoid type belief networks, a class of probabilistic neural networks, provide a natural framework for compactly representing probabilistic information in a variety of unsupervised and supervised...
Dimensionality Reduction for Supervised Learning With Reproducing Kernel Hilbert Spaces (1998)
Fukumizu, Kenji, Bach, Francis R., Jordan, Michael I.
We propose a novel method of dimensionality reduction for supervised learning problems. Given a regression or classifcation problem in which we wish to predict a response variable Y from an...
Decentralized Detection and Classification Using Kernel Methods (1998)
Nguyen, XuanLong, Wainwright, Martin J., Jordan, Michael I.
We consider the problem of decentralized detection under constraints on the number of bits that can be transmitted by each sensor. In contrast to most previous work, in which the joint We consider...
Estimating Dependency Structure as a Hidden Variable (1998)
Meila, Marina, Jordan, Michael I., Morris, Quaid
This paper introduces a probability model, the mixture of trees that can account for sparse, dynamically changing dependence relationships. We present a family of efficient algorithms that use EM and...
Estimating Dependency Structure as a Hidden Variable (1998)
Meila, Marina, Jordan, Michael I., Morris, Quaid
This paper introduces a probability model, the mixture of trees that can account for sparse, dynamically changing dependence relationships. We present a family of efficient algorithms that use EM and...
Estimating dependency structure as a hidden variable (1998)
Marina Meilă, Michael I. Jordan, Quaid Morris
This publication can be retrieved by anonymous ftp to publications.ai.mit.edu. This paper introduces a probability model, the mixture of trees that can account for sparse, dynamically changing...
Learning in graphical models (1998)
Statistical applications in fields such as bioinformatics, information retrieval, speech processing, im-age processing and communications often involve large-scale models in which thousands or...
Learning in graphical models (1998)
Statistical applications in fields such as bioinformatics, information retrieval, speech processing, im-age processing and communications often involve large-scale models in which thousands or...
Learning in graphical models (1998)
Statistical applications in fields such as bioinformatics, information retrieval, speech processing, im-age processing and communications often involve large-scale models in which thousands or...
Learning in graphical models (1998)
Statistical applications in fields such as bioinformatics, information retrieval, speech processing, im-age processing and communications often involve large-scale models in which thousands or...
Estimating dependency structure as a hidden variable (1998)
Marina Meila, Michael I. Jordan, Quaid Morris
This publication can be retrieved by anonymous ftp to publications.ai.mit.edu. This paper introduces a probability model,themixture oftrees that can account for sparse, dynamically changing...
Estimating dependency structure as a hidden variable (1998)
Marina Meila, Michael I. Jordan, Quaid Morris
This publication can be retrieved by anonymous ftp to publications.ai.mit.edu. This paper introduces a probability model, the mixture of trees that can account for sparse, dynamically changing...
An Introduction to Variational Methods for Graphical Methods (1998)
Michael I. Jordan, Zoubin Ghahramani, Tommi S. Jaakkola, Lawrence K. Saul
. This paper presents a tutorial introduction to the use of variational methods for inference and learning in graphical models (Bayesian networks and Markov random fields). We present a number of...
The Role of Inertial Sensitivity in Motor Planning (1998)
Philip N. Sabes, Michael I. Jordan, Daniel M. Wolpert
This paper addresses a number of unresolved issues.
Improving the Mean Field Approximation via the Use of Mixture Distributions (1998)
Tommi S. Jaakkola, Michael I. Jordan
. Mean field methods provide computationally efficient approximations to posterior probability distributions for graphical models. Simple mean field methods make a completely factorized approximation...
Lawrence K. Saul, Michael I. Jordan
. We study Markov models whose state spaces arise from the Cartesian product of two or more discrete random variables. We show how to parameterize the transition matrices of these models as a convex...
Estimating Dependency Structure as a Hidden Variable (1998)
Marina Meila, Michael I. Jordan
This paper introduces a probability model, the mixture of trees that can account for sparse, dynamically changing dependence relationships. We present a family of efficient algorithms that use EM and...
The Role of Inertial Sensitivity in Motor Planning (1998)
Behavioral/systems Neuroscience Section, Philip N. Sabes, Michael I. Jordan, Daniel M. Wolpert
, 252; Introduction, 650; Discussion, 1280 Acknowledgments: We thank N. Hogan for many helpful discussions. This project was supported by grants from the Wellcome Trust and the U.S. Office of Naval...
Variational probabilistic inference and the QMR-DT database (1998)
Tommi S. Jaakkola, Michael I. Jordan
We describe a variational approximation method for efficient inference in large-scale probabilistic models. Variational methods are deterministic procedures that provide approximations to marginal...
Bayesian Parameter Estimation Through Variational Methods (1998)
Tommi S. Jaakkola, Michael I. Jordan
We consider a logistic regression model with a Gaussian prior distribution over the parameters. We show that accurate variational techniques can be used to obtain a closed form posterior distribution...
Mixture Representations for Inference and Learning in Boltzmann Machines (1998)
Neil D. Lawrence, Christopher M. Bishop, Michael I. Jordan
Boltzmann machines are undirected graphical models with two-state stochastic variables, in which the logarithms of the clique potentials are quadratic functions of the node states. They have been...
On the Concentration of Expectation and (1998)
Approximate Inference In, Xuanlong Nguyen, Michael I. Jordan
We present an analysis of concentration-of-expectation phenomena in layered Bayesian networks that use generalized linear models as the local conditional probabilities. This framework encompasses a...
Validation and Abstraction (1998)
Xuanlong Nguyen, Michael I. Jordan
approximate inference in layered networks
Validation and Abstraction (1998)
Michael I. Jordan, Laurent El Ghaoui
PCA using semidefinite programming
Approximating Posterior Distributions in Belief Networks using Mixtures (1998)
Christopher M. Bishop, Neil D. Lawrence, Tommi Jaakkola, Michael I. Jordan
Exact inference in densely connected Bayesian networks is computationally intractable, and so there is considerable interest in developing effective approximation schemes. One approach which has been...
Estimating Dependency Structure as a Hidden Variable (1997)
Meila, Marina, Jordan, Michael I., Morris, Quaid
This paper introduces a probability model, the mixture of trees that can account for sparse, dynamically changing dependence relationships. We present a family of efficient algorithms that use EMand...
Estimating Dependency Structure as a Hidden Variable (1997)
Meila, Marina, Jordan, Michael I., Morris, Quaid
This paper introduces a probability model, the mixture of trees that can account for sparse, dynamically changing dependence relationships. We present a family of efficient algorithms that use EMand...
Triangulation by Continuous Embedding (1997)
Meila, Marina, Jordan, Michael I.
When triangulating a belief network we aim to obtain a junction tree of minimum state space. Searching for the optimal triangulation can be cast as a search over all the permutations of the network's...
Triangulation by Continuous Embedding (1997)
Meila, Marina, Jordan, Michael I.
When triangulating a belief network we aim to obtain a junction tree of minimum state space. Searching for the optimal triangulation can be cast as a search over all the permutations of the network's...
Probabilistic independence networks for hidden Markov probability models (1997)
Smyth, Padhraic, Heckerman, David, Jordan, Michael I.
Graphical techniques for modeling the dependencies of random variables have been explored in a variety of different areas, including statistics, statistical physics, artificial intelligence, speech...
Factorial Hidden Markov Models. (1997)
Ghahramani, Zoubin, Jordan, Michael I.
We present a framework for learning in hidden Markov models with distributed state representations. Within this framework, we derive a learning algorithm based on the Expectation-Maximization (EM)...
Rationality and intelligence (1997)
Michael I. Jordan, Stuart Russell
There are two complementary views of artificial intelligence (AI): one as an engineering discipline concerned with the creation of intelligent machines, the other as an empirical science concerned...
Obstacle avoidance and a perturbation sensitivity model for motor planning (1997)
Philip N. Sabes, Michael I. Jordan
A novel obstacle avoidance paradigm was used to investigate the planning of human reaching movements. We explored whether the CNS plans arm movements based entirely on the visual space kinematics of...
Factorial hidden Markov models (1997)
Zoubin Ghahramani, Michael I. Jordan, Padhraic Smyth
Abstract. Hidden Markov models (HMMs) have proven to be one of the most widely used tools for learning probabilistic models of time series data. In an HMM, information about the past is conveyed...
Triangulation by Continuous Embedding (1997)
Marina Meila, Michael I. Jordan
When triangulating a belief network we aim to obtain a junction tree of minimum state space. According to (Rose, 1970), searching for the optimal triangulation can be cast as a search over all the...
Hidden Markov decision trees (1997)
Michael I. Jordan, Zoubin Ghahramani, Lawrence K. Saul
We study a time series model that can be viewed as a decision tree with Markov temporal structure. The model is intractable for exact calculations, thus we utilize variational approximations. We...
Computational Models of Sensorimotor Integration (1997)
Zoubin Ghahramani Daniel, Daniel M. Wolpert, Michael I. Jordan
The sensorimotor integration system can be viewed as an observer attempting to estimate its own state and the state of the environment by integrating multiple sources of information. We describe a...
Triangulation by Continuous Embedding (1997)
Marina Meila, Michael I. Jordan
When triangulating a belief network we aim to obtain a junction tree of minimumstate space. According to [8], searching for the optimal triangulation can be cast as a search over all the permutations...
Obstacle Avoidance and a Perturbation Sensitivity Model for Motor Planning (1997)
Michael I. Jordan, Philip N. Sabes, Philip N. Sabes
A novel obstacle avoidance paradigm was used to investigate the planning of human reaching movements. We explored whether the central nervous system (CNS) plans arm movements based entirely on the...
Estimating Dependency Structure as a Hidden Variable (1997)
Marina Meila, Michael I. Jordan, Quaid Morris
This paper introduces a probability model, the mixture of trees that can account for sparse, dynamically changing dependence relationships. We present a family of efficient algorithms that use EM and...
Factorial Hidden Markov Models (1997)
Zoubin Ghahramani, Michael I. Jordan, Padhraic Smyth
. Hidden Markov models (HMMs) have proven to be one of the most widely used tools for learning probabilistic models of time series data. In an HMM, information about the past is conveyed through a...
Computational Models of Sensorimotor Integration (1997)
Zoubin Ghahramani, Daniel M. Wolpert, Michael I. Jordan
The sensorimotor integration system can be viewed as an observer attempting to estimate its own state and the state of the environment by integrating multiple sources of information. We describe a...
Mean Field Theory for Sigmoid Belief Networks (1996)
Saul, Lawrence K., Jaakkola, Tommi, Jordan, Michael I.
We develop a mean field theory for sigmoid belief networks based on ideas from statistical mechanics. Our mean field theory provides a tractable approximation to the true probability distribution in...
Mean Field Theory for Sigmoid Belief Networks (1996)
Saul, Lawrence K., Jaakkola, Tommi, Jordan, Michael I.
We develop a mean field theory for sigmoid belief networks based on ideas from statistical mechanics. Our mean field theory provides a tractable approximation to the true probability distribution in...
Jordan, Michael I., Bishop, Christopher M.
We present an overview of current research on artificial neural networks, emphasizing a statistical perspective. We view neural networks as parameterized graphs that make probabilistic assumptions...
Jordan, Michael I., Bishop, Christopher M.
We present an overview of current research on artificial neural networks, emphasizing a statistical perspective. We view neural networks as parameterized graphs that make probabilistic assumptions...
Computing Upper and Lower Bounds on Likelihoods in Intractable Networks (1996)
Jaakkola, Tommi S., Jordan, Michael I.
We present techniques for computing upper and lower bounds on the likelihoods of partial instantiations of variables in sigmoid and noisy-OR networks. The bounds determine confidence intervals for...
Computing Upper and Lower Bounds on Likelihoods in Intractable Networks (1996)
Jaakkola, Tommi S., Jordan, Michael I.
We present techniques for computing upper and lower bounds on the likelihoods of partial instantiations of variables in sigmoid and noisy-OR networks. The bounds determine confidence intervals for...
Factorial Hidden Markov Models (1996)
Ghahramani, Zoubin, Jordan, Michael I.
We present a framework for learning in hidden Markov models with distributed state representations. Within this framework, we derive a learning algorithm based on the Expectation--Maximization (EM)...
Fast Learning by Bounding Likelihoods in Sigmoid Type Belief Networks (1996)
Jaakkola, Tommi S., Saul, Lawrence K., Jordan, Michael I.
Sigmoid type belief networks, a class of probabilistic neural networks, provide a natural framework for compactly representing probabilistic information in a variety of unsupervised and supervised...
Fast Learning by Bounding Likelihoods in Sigmoid Type Belief Networks (1996)
Jaakkola, Tommi S., Saul, Lawrence K., Jordan, Michael I.
Sigmoid type belief networks, a class of probabilistic neural networks, provide a natural framework for compactly representing probabilistic information in a variety of unsupervised and supervised...
Factorial Hidden Markov Models (1996)
Ghahramani, Zoubin, Jordan, Michael I.
We present a framework for learning in hidden Markov models with distributed state representations. Within this framework, we derive a learning algorithm based on the Expectation--Maximization (EM)...
Michael I. Jordan, Christopher M. Bishop
Neural networks have emerged as a field
Learning fine motion by Markov mixtures of experts (1996)
Marina Meila, Michael I. Jordan
This publication can be retrieved by anonymous ftp to publications.ai.mit.edu. Compliantcontrol is a standard method for performing ne manipulation tasks, like grasping and assembly, but it requires...
On Convergence Properties of the EM Algorithm for Gaussian Mixtures (1996)
This publication can be retrieved by anonymous ftp to publications.ai.mit.edu. We build up the mathematical connection between the \Expectation-Maximization " (EM) algorithm and...
Fast learning by bounding likelihoods in sigmoid type belief networks (1996)
Tommi S. Jaakkola, Lawrence K. Saul, Michael I. Jordan
This publication can be retrieved by anonymous ftp to publications.ai.mit.edu. Sigmoid type belief networks, a class of probabilistic neural networks, provide a natural framework for compactly...
Reinforcement learning by probability matching (1996)
Philip N. Sabes, Michael I. Jordan
This publication can be retrieved by anonymous ftp to publications.ai.mit.edu. We present a new algorithm for associative reinforcement learning. The algorithm is based upon the idea of matching a...
Reinforcement learning by probability matching (1996)
Philip N. Sabes, Michael I. Jordan
This publication can be retrieved by anonymous ftp to publications.ai.mit.edu. We present a new algorithm for associative reinforcement learning. The algorithm is based upon the idea of matching a...
Mean field theory for sigmoid belief networks (1996)
Lawrence K. Saul, Tommi Jaakkola, Michael I. Jordan
This publication can be retrieved by anonymous ftp to publications.ai.mit.edu. We develop a mean field theory for sigmoid belief networks based on ideas from statistical mechanics. Our mean field...
On Convergence Properties of the EM Algorithm for Gaussian Mixtures (1996)
This publication can be retrieved by anonymous ftp to publications.ai.mit.edu. We build up the mathematical connection between the "Expectation-Maximization " (EM) algorithm and...
Fast learning by bounding likelihoods in sigmoid type belief networks (1996)
Tommi Jaakkola, Lawrence K. Saul, Michael I. Jordan
Sigmoid type belief networks, a class of probabilistic neural networks, provide a natural framework for compactly representing probabilistic information in a variety of unsupervised and supervised...
Mean field theory for sigmoid belief networks (1996)
Lawrence K. Saul, Tommi Jaakkola, Michael I. Jordan
We develop a mean field theory for sigmoid belief networks based on ideas from statistical mechanics. Our mean field theory provides a tractable approximation to the true probability distribution in...
Michael I. Jordan, Christopher M. Bishop
We present an overview of current research on artificial neural networks, emphasizing a statistical perspective. We view neural networks as parameterized graphs that make probabilistic assumptions...
On Convergence Properties of the EM Algorithm for Gaussian Mixtures (1996)
This publication can be retrieved by anonymous ftp to publications.ai.mit.edu. We build up the mathematical connection between the "Expectation-Maximization " (EM) algorithm and...
Computing Upper and Lower Bounds on Likelihoods in Intractable Networks (1996)
Tommi S. Jaakkola, Michael I. Jordan
We present deterministic techniques for computing upper and lower bounds on marginal probabilities in sigmoid and noisy-OR networks. These techniques become useful when the size of the network (or...
Computational Aspects of Motor Control and Motor Learning (1996)
This chapter provides a basic introduction to various of the computational issues that arise in the study of motor control and motor learning. A broad set of topics is discussed, including feedback...
A Variational Principle for Model-Based Interpolation (1996)
Lawrence Saul, Michael I. Jordan
Given a multidimensional data set and a model of its density, we consider how to define the optimal interpolation between two points. This is done by assigning a cost to each path through space,...
Local Linear Perceptrons for Classification (1996)
Ethem Alpaydin, Michael I Jordan
A structure composed of local linear perceptrons for approximating global class discriminants is investigated. Such local linear models may be combined in a cooperative or competitive way. In the...
Active Learning with Statistical Models (1996)
David A. Cohn, Zoubin Ghahramani, Michael I. Jordan
For manytypes of machine learning algorithms, one can compute the statistically "optimal" way to select training data. In this paper, we review how optimal data selection techniques have...
Generalization to Local Remappings of the Visuomotor Coordinate Transformation (1996)
Zoubin Ghahramani Daniel, Daniel M. Wolpert, Michael I. Jordan
this paper, we explore the representation of the visuomotor transformation by analyzing the pattern of adaptation arising from a local perturbation of the visuomotor relationship. We use a modern-day...
Computing Upper and Lower Bounds on Likelihoods in Intractable Networks (1996)
Tommi Jaakkola And, Tommi S. Jaakkola, Michael I. Jordan
We present techniques for computing upper and lower bounds on the likelihoods of partial instantiations of variables in sigmoid and noisy-OR networks. The bounds determine confidence intervals for...
Factorial Hidden Markov Models (1996)
Zoubin Ghahramani Zoubin, Michael I. Jordan
We present a framework for learning in hidden Markov models with distributed state representations. Within this framework, we derive a learning algorithm based on the Expectation--Maximization (EM)...
Probabilistic Independence Networks for Hidden Markov Probability Models (1996)
Padhraic Smyth, David Heckerman, Michael I. Jordan
Graphical techniques for modeling the dependencies of random variables have been explored in a variety of different areas including statistics, statistical physics, artificial intelligence, speech...
Computing Upper and Lower Bounds on Likelihoods in Intractable Networks (1996)
Tommi Jaakkola, Michael I. Jordan
We present techniques for computing upper and lower bounds on the likelihoods of partial instantiations of variables in sigmoid and noisy-OR networks. The bounds determine confidence intervals for...
Reinforcement Learning by Probability Matching (1996)
Philip Sabes, Michael I. Jordan
We present a new algorithm for associative reinforcement learning. The algorithm is based upon the idea of matching a network's output probability with a probability distribution derived from...
Probabilistic Independence Networks for Hidden Markov Probability Models (1996)
Padhraic Smyth, David Heckerman, Michael I. Jordan
Graphical techniques for modeling the dependencies of random variables have been explored in a variety of different areas including statistics, statistical physics, artificial intelligence, speech...
A variational approach to Bayesian logistic regression models and their extensions (1996)
Tommi S. Jaakkola, Michael I. Jordan
We consider a logistic regression model with a Gaussian prior distribution over the parameters. We show that accurate variational techniques can be used to obtain a closed form posterior distribution...
Factorial Hidden Markov Models (1996)
Zoubin Ghahramani, Michael I. Jordan
We present a framework for learning in hidden Markov models with distributed state representations. Within this framework, we derive a learning algorithm based on the Expectation--Maximization (EM)...
Generalization to Local Remappings of the Visuomotor Coordinate Transformation (1996)
Zoubin Ghahramani, Daniel M. Wolpert, Michael I. Jordan
To investigate the representation and plasticity of the visuomotor coordinate transformation we examined the changes in pointing behavior subsequent to local remappings. The visual feedback of finger...
Active Learning with Statistical Models (1996)
David A. Cohn, Zoubin Ghahramani, Michael I. Jordan
For many types of machine learning algorithms, one can compute the statistically "optimal " way to select training data. In this paper, we review how optimal data selection techniques have...
Reinforcement Learning by Probability Matching (1996)
Philip Sabes, Michael I. Jordan
We present a new algorithm for associative reinforcement learning. The algorithm is based upon the idea of matching a network's output probability with a probability distribution derived from...
On Convergence Properties of the EM Algorithm for Gaussian Mixtures (1996)
We build up the mathematical connection between the "Expectation-Maximization" (EM) algorithm and gradient-based approaches for maximum likelihood learning of finite Gaussian mixtures. We...
Fast Learning by Bounding Likelihoods in Sigmoid Type Belief Networks (1996)
Tommi S. Jaakkola, Lawrence K. Saul, Michael I. Jordan
Sigmoid type belief networks, a class of probabilistic neural networks, provide a natural framework for compactly representing probabilistic information in a variety of unsupervised and supervised...
Mean Field Theory for Sigmoid Belief Networks (1996)
Lawrence Saul, Tommi Jaakkola, Michael I. Jordan
In this paper we show how to calculate a rigorous lower bound on the likelihood of observed activities in sigmoid belief networks. We view these networks in the framework of statistical mechanics and...
Factorial Hidden Markov Models (1996)
Zoubin Ghahramani, Michael I. Jordan
One of the basic probabilistic tools used for time series modeling is the hidden Markov model (HMM). In an HMM, information about the past of the time series is conveyed through a single discrete...
Learning Fine Motion by Markov Mixtures of Experts (1996)
Marina Meila, Michael I. Jordan
Compliantcontrol is a standard method for performing fine manipulation tasks, like grasping and assembly, but it requires estimation of the state of contact between the robot arm and the objects...
Mean Field Theory for Sigmoid Belief Networks (1996)
Lawrence K. Saul, Tommi Jaakkola, Michael I. Jordan
We develop a mean field theory for sigmoid belief networks based on ideas from statistical mechanics. Our mean field theory provides a tractable approximation to the true probability distribution in...
Mean Field Theory for Sigmoid Belief Networks (1996)
Lawrence K. Saul, Tommi Jaakkola, Michael I. Jordan
We develop a mean field theory for sigmoid belief networks based on ideas from statistical mechanics. Our mean field theory provides a tractable approximation to the true probability distribution in...
Generalization to Local Remappings of the Visuomotor Coordinate Transformation (1996)
Zoubin Ghahramani, Daniel M. Wolpert, Michael I. Jordan
this article should be addressed to: Zoubin Ghahramani, Department of Computer Science, University of Toronto, 6 King's College Rd. Pratt 271, Toronto, CANADA M5S 3H5. Tel: (416) 978-7453, Fax:...
Mean field theory for sigmoid belief networks (1996)
Lawrence K. Saul, Tommi Jaakkola, Michael I. Jordan
We develop a mean eld theory for sigmoid belief networks based on ideas from statistical mechanics. Our mean eld theory provides a tractable approximation to the true probability distribution in...
Learning Fine Motion by Markov Mixtures of Experts (1995)
Meila, Marina, Jordan, Michael I.
Compliant control is a standard method for performing fine manipulation tasks, like grasping and assembly, but it requires estimation of the state of contact between the robot arm and the objects...
Learning Fine Motion by Markov Mixtures of Experts (1995)
Meila, Marina, Jordan, Michael I.
Compliant control is a standard method for performing fine manipulation tasks, like grasping and assembly, but it requires estimation of the state of contact between the robot arm and the objects...
Active Learning with Statistical Models (1995)
Cohn, David A., Ghahramani, Zoubin, Jordan, Michael I.
For many types of learners one can compute the statistically 'optimal' way to select data. We review how these techniques have been used with feedforward neural networks. We then show how the same...
Active Learning with Statistical Models (1995)
Cohn, David A., Ghahramani, Zoubin, Jordan, Michael I.
For many types of learners one can compute the statistically 'optimal' way to select data. We review how these techniques have been used with feedforward neural networks. We then show how the same...
Learning from Incomplete Data (1995)
Ghahramani, Zoubin, Jordan, Michael I.
Real-world learning tasks often involve high-dimensional data sets with complex patterns of missing features. In this paper we review the problem of learning from incomplete data from two statistical...
Learning from Incomplete Data (1995)
Ghahramani, Zoubin, Jordan, Michael I.
Real-world learning tasks often involve high-dimensional data sets with complex patterns of missing features. In this paper we review the problem of learning from incomplete data from two statistical...
Computational structure of coordinate transformations: a generalization study (1995)
Zoubin Ghahramani, Daniel M. Wolpert, Michael I. Jordan
One of the fundamental properties that both neural networks and the central nervous system share is the ability to learn and generalize from examples. While this property has been studied extensively...
Convergence results for the EM approach to mixtures of experts architectures (1995)
The Expectation-Maximization (EM) algorithm is an iterative approach to maximum likelihood parameter estimation. Jordan and Jacobs (1993) recently proposed an EM algorithm for the mixture of experts...
Reinforcement Learning with Soft State Aggregation (1995)
Satinder P. Singh, Tommi Jaakkola, Michael I. Jordan
It is widely accepted that the use of more compact representations than lookup tables is crucial to scaling reinforcement learning (RL) algorithms to real-world problems. Unfortunately almost all of...
Reinforcement Learning Algorithm for Partially Observable Markov Decision Problems (1995)
Tommi Jaakkola, Satinder P. Singh, Michael I. Jordan
Increasing attention has been paid to reinforcement learning algorithms in recent years, partly due to successes in the theoretical analysis of their behavior in Markov environments. If the Markov...
Reinforcement Learning with Soft State Aggregation (1995)
Satinder Singh Singh, Tommi Jaakkola, Michael I. Jordan
It is widely accepted that the use of more compact representations than lookup tables is crucial to scaling reinforcement learning (RL) algorithms to real-world problems. Unfortunately almost all of...
Active Learning with Statistical Models (1995)
David A. Cohn, Zoubin Ghahramani, Michael I. Jordan
For many types of learners one can compute the statistically "optimal" way to select data. We review how these techniques have been used with feedforward neural networks [MacKay, 1992;...
Are arm trajectories planned in kinematic or dynamic coordinates? An adaptation study (1995)
Daniel Wolpert, Zoubin Ghahramani, Michael I. Jordan
There are several invariant features of point-to-point human arm movements: trajectories tend to be straight, smooth and have bell-shaped velocity profiles. One approach to accounting for these data...
Exploiting Tractable Substructures in Intractable Networks (1995)
Lawrence Saul, Michael I. Jordan
We develop a refined mean field approximation for inference and learning in probabilistic neural networks. Our mean field theory, unlike most, does not assume that the units behave as independent...
Factorial hidden Markov models (1995)
Zoubin Ghahramani, Michael I. Jordan
We present a framework for learning in hidden Markov models with distributed state representations. Within this framework, we derive a learning algorithm based on the Expectation--Maximization (EM)...
Boltzmann Chains and Hidden Markov Models (1995)
Lawrence K. Saul, Michael I. Jordan
We propose a statistical mechanical framework for the modeling of discrete time series. Maximum likelihood estimation is done via Boltzmann learning in one-dimensional networks with tied weights. We...
Computational structure of coordinate transformations: A generalization study (1995)
Zoubin Ghahramani, Daniel M. Wolpert, Michael I. Jordan
One of the fundamental properties that both neural networks and the central nervous system share is the ability to learn and generalize from examples. While this property has been studied extensively...
Active Learning with Statistical Models (1995)
David A. Cohn, Zoubin Ghahramani, Michael I. Jordan
For many types of learners one can compute the statistically "optimal " way to select data. We review how these techniques have been used with feedforward neural networks [MacKay, 1992;...
Learning From Incomplete Data (1995)
Zoubin Ghahramani, Michael I. Jordan
Real-world learning tasks often involve high-dimensional data sets with complex patterns of missing features. In this paper we review the problem of learning from incomplete data from two statistical...
An Alternative Model for Mixtures of Experts (1995)
Lei Xu, Michael I. Jordan, Geoffrey E. Hinton
An alternative model is proposed for mixtures-of-experts, by utilizing a different parametric form for the gating network. The modified model is trained byanEM algorithm. In comparison with earlier...
Reinforcement learning algorithm for partially observable markov decision problems (1995)
Tommi Jaakkola, Satinder P. Singh, Michael I. Jordan
Increasing attention has been paid to reinforcement learning algorithms in recent years, partly due to successes in the theoretical analysis of their behaviorinMarkov environments. If the Markov...
Advances in neural information processing systems (1995)
Philip N. Sabes, Michael I. Jordan
We present a new algorithm for associative reinforcement learning. The algorithm is based upon the idea of matching a network's output probability with a probability distribution derived from...
Hierarchical mixtures of experts and the EM algorithm (1994)
We present a tree-structured architecture for supervised learning. The statistical model underlying the architecture is a hi-erarchical mixture model in which both the mixture coefficients and the...
Hierarchical mixtures of experts and the EM algorithm (1994)
Michael I. Jordan, Robert A. Jacobs
We present a tree-structured architecture for supervised learning. The statistical model underlying the architecture is a hierarchical mixture model in which both the mixture coe cients and the...
On the convergence of stochastic iterative dynamic programming algorithms (1994)
Tommi Jaakkola, Michael I. Jordan, Satinder P. Singh
Increasing attention has recently been paid to algorithms based on dynamic programming (DP) due to the suitability of DP for learning problems involving control. In stochastic environments where the...
Learning from incomplete data (1994)
Zoubin Ghahramani, Michael I. Jordan
This publication can be retrieved by anonymous ftp to publications.ai.mit.edu. Real-world learning tasks often involve high-dimensional data sets with complex patterns of missing features. In this...
On the convergence of stochastic iterative dynamic programming algorithms (1994)
Tommi Jaakkola, Michael I. Jordan, Satinder P. Singh
Recent developments in the area of reinforcement learning have yielded a number of new algorithms for the prediction and control of Markovian environments. These algorithms, including the TD()...
On the Convergence of Stochastic Iterative Dynamic Programming Algorithms (1994)
Tommi Jaakkola, Michael I. Jordan, Satinder P. Singh
Recent developments in the area of reinforcement learning have yielded a number of new algorithms for the prediction and control of Markovian environments. These algorithms, including the TD()...
A Statistical Approach to Decision Tree Modeling (1994)
A statistical approach to decision tree modeling is described. In this approach, each decision in the tree is modeled parametrically as is the process by which an output is generated from an input...
On the Convergence of Stochastic Iterative Dynamic Programming Algorithms (1994)
Tommi Jaakkola, Michael I. Jordan, Satinder P. Singh
Recent developments in the area of reinforcement learning have yielded a number of new algorithms for the prediction and control of Markovian environments. These algorithms, including the TD()...
Learning Without State-Estimation in Partially Observable Markovian Decision Processes (1994)
Satinder P. Singh, Tommi Jaakkola, Michael I. Jordan
Reinforcement learning (RL) algorithms provide a sound theoretical basis for building learning control architectures for embedded agents. Unfortunately all of the theory and much of the practice (see...
Hierarchical mixtures of experts and the EM algorithm (1994)
Michael I. Jordan, Robert A. Jacobs
We present a tree-structured architecture for supervised learning. The statistical model underlying the architecture is a hierarchical mixture model in which both the mixture coefficients and the...
Perceptual Distortion Contributes to the Curvature of Human Reaching Movements. (1994)
Daniel Wolpert, Zoubin Ghahramani, Michael I. Jordan
Unconstrained point-to-point human arm movements are generally gently curved, a fact which has been used to assess the validity of models of trajectory formation. In this study we examined the...
Supervised learning from incomplete data via an EM approach (1994)
Zoubin Ghahramani, Michael I. Jordan
Real-world learning tasks may involve high-dimensional data sets with arbitrary patterns of missing data. In this paper we present a framework based on maximum likelihood density estimation for...
On the convergence of stochastic iterative dynamic programming algorithms (1994)
Tommi Jaakkola, Michael I. Jordan, Satinder P. Singh
putational Learning at MIT, including funds provided by DARPA under the
Convergence Results for the EM Approach to Mixtures of Experts Architectures (1993)
The Expectation-Maximization (EM) algorithm is an iterative approach to maximum likelihood parameter estimation. Jordan and Jacobs (1993) recently proposed an EM algorithm for the mixture of experts...
Convergence Results for the EM Approach to Mixtures of Experts Architectures (1993)
The Expectation-Maximization (EM) algorithm is an iterative approach to maximum likelihood parameter estimation. Jordan and Jacobs (1993) recently proposed an EM algorithm for the mixture of experts...
On the Convergence of Stochastic Iterative Dynamic Programming Algorithms (1993)
Jaakkola, Tommi, Jordan, Michael I., Singh, Satinder P.
Recent developments in the area of reinforcement learning have yielded a number of new algorithms for the prediction and control of Markovian environments. These algorithms, including the TD(lambda)...
Hierarchical Mixtures of Experts and the EM Algorithm (1993)
Jordan, Michael I., Jacobs, Robert A.
We present a tree-structured architecture for supervised learning. The statistical model underlying the architecture is a hierarchical mixture model in which both the mixture coefficients and the...
Hierarchical Mixtures of Experts and the EM Algorithm (1993)
Jordan, Michael I., Jacobs, Robert A.
We present a tree-structured architecture for supervised learning. The statistical model underlying the architecture is a hierarchical mixture model in which both the mixture coefficients and the...
On the Convergence of Stochastic Iterative Dynamic Programming Algorithms (1993)
Jaakkola, Tommi, Jordan, Michael I., Singh, Satinder P.
Recent developments in the area of reinforcement learning have yielded a number of new algorithms for the prediction and control of Markovian environments. These algorithms, including the TD(lambda)...
Convergence results for the EM approach to mixtures of experts architectures (1993)
The Expectation-Maximization (EM) algorithm is an iterative approach to maximum likelihood parameter estimation. Jordan and Jacobs (1993) recently proposed an EM algorithm for the mixture of experts...
Forward models: Supervised learning with a distal teacher (1992)
Michael I. Jordan, David E. Rumelhart
Internal models of the environment have an important role to play in adaptive systems in general and are of particular importance for the supervised learning paradigm. In this paper we demonstrate...
Reinforcement Learning And Its Application To Control (1992)
Vijaykumar Gullapalli, Roderic A. Grupen, B. Erik Ydstie, Michael I. Jordan, W. Richards Adrion
REINFORCEMENT LEARNING AND ITS APPLICATION TO CONTROL February 1992 Vijaykumar Gullapalli, B.S., Birla Institute of Technology and Science, India M.S., University of Massachusetts Ph.D., University...
Reinforcement Learning And Its Application To Control (1992)
Vijaykumar Gullapalli, Roderic A. Grupen, B. Erik Ydstie, Michael I. Jordan, W. Richards Adrion
REINFORCEMENT LEARNING AND ITS APPLICATION TO CONTROL February 1992 Vijaykumar Gullapalli, B.S., Birla Institute of Technology and Science, India M.S., University of Massachusetts Ph.D., University...
Robert A. Jacobs, A. Jacobs, Michael I. Jordan
The influence of Andrew Barto and Michael Jordan on my thinking is evidenced throughout this dissertation. They provided technical training, constructive criticism, guidance, and inspiration...
Mazzoni, Pietro, Andersen, Richard A., Jordan, Michael I.
Area 7a of the posterior parietal cortex of the primate brain is concerned with representing head-centered space by combining information about the retinal location of a visual stimulus and the...
Robert A. Jacobs, Michael I. Jordan, Andrew G. Barto
A novel modular connectionist architecture is presented in which the networks composing the architecture compete to learn the training patterns. An outcome of the competition is that different...
Toward a protein profile of Escherichia coli: Comparison to its transcription profile
Corbin, Rebecca W., Paliy, Oleg, Yang, Feng, Shabanowitz, Jeffrey, Platt, Mark, Lyons, Charles E., ...
High-pressure liquid chromatography–tandem mass spectrometry was used to obtain a protein profile of Escherichia coli strain MG1655 grown in minimal medium with glycerol as the carbon source. By...
Chemogenomic profiling: Identifying the functional interactions of small molecules in yeast
Giaever, Guri, Flaherty, Patrick, Kumm, Jochen, Proctor, Michael, Nislow, Corey, Jaramillo, Daniel F., ...
We demonstrate the efficacy of a genome-wide protocol in yeast that allows the identification of those gene products that functionally interact with small molecules and result in the inhibition of...
Sulfur and Nitrogen Limitation in Escherichia coli K-12: Specific Homeostatic Responses
Gyaneshwar, Prasad, Paliy, Oleg, McAuliffe, Jon, Popham, David L., Jordan, Michael I., Kustu, Sydney
We determined global transcriptional responses of Escherichia coli K-12 to sulfur (S)- or nitrogen (N)-limited growth in adapted batch cultures and cultures subjected to nutrient shifts. Using two...
Gyaneshwar, Prasad, Paliy, Oleg, McAuliffe, Jon, Jones, Adriane, Jordan, Michael I., Kustu, Sydney
We previously characterized nutrient-specific transcriptional changes in Escherichia coli upon limitation of nitrogen (N) or sulfur (S). These global homeostatic responses presumably minimize the...
Subtree power analysis and species selection for comparative genomics
McAuliffe, Jon D., Jordan, Michael I., Pachter, Lior
Sequence comparison across multiple organisms aids in the detection of regions under selection. However, resource limitations require a prioritization of genomes to be sequenced. This prioritization...
Genome-Wide Requirements for Resistance to Functionally Distinct DNA-Damaging Agents
Lee, William, St.Onge, Robert P, Proctor, Michael, Flaherty, Patrick, Jordan, Michael I, Arkin, Adam P, ...
The mechanistic and therapeutic differences in the cellular response to DNA-damaging compounds are not completely understood, despite intense study. To expand our knowledge of DNA damage, we assayed...
Protein Molecular Function Prediction by Bayesian Phylogenomics
Engelhardt, Barbara E, Jordan, Michael I, Muratore, Kathryn E, Brenner, Steven E
We present a statistical graphical model to infer specific molecular function for unannotated protein sequences using homology. Based on phylogenomic principles, SIFTER (Statistical Inference of...
Toward a protein profile of Escherichia coli: Comparison to its transcription profile
Corbin, Rebecca W., Paliy, Oleg, Yang, Feng, Shabanowitz, Jeffrey, Platt, Mark, Lyons, Charles E., ...
High-pressure liquid chromatography–tandem mass spectrometry was used to obtain a protein profile of Escherichia coli strain MG1655 grown in minimal medium with glycerol as the carbon source. By...
Chemogenomic profiling: Identifying the functional interactions of small molecules in yeast
Giaever, Guri, Flaherty, Patrick, Kumm, Jochen, Proctor, Michael, Nislow, Corey, Jaramillo, Daniel F., ...
We demonstrate the efficacy of a genome-wide protocol in yeast that allows the identification of those gene products that functionally interact with small molecules and result in the inhibition of...
Sulfur and Nitrogen Limitation in Escherichia coli K-12: Specific Homeostatic Responses
Gyaneshwar, Prasad, Paliy, Oleg, McAuliffe, Jon, Popham, David L., Jordan, Michael I., Kustu, Sydney
We determined global transcriptional responses of Escherichia coli K-12 to sulfur (S)- or nitrogen (N)-limited growth in adapted batch cultures and cultures subjected to nutrient shifts. Using two...
Gyaneshwar, Prasad, Paliy, Oleg, McAuliffe, Jon, Jones, Adriane, Jordan, Michael I., Kustu, Sydney
We previously characterized nutrient-specific transcriptional changes in Escherichia coli upon limitation of nitrogen (N) or sulfur (S). These global homeostatic responses presumably minimize the...
Subtree power analysis and species selection for comparative genomics
McAuliffe, Jon D., Jordan, Michael I., Pachter, Lior
Sequence comparison across multiple organisms aids in the detection of regions under selection. However, resource limitations require a prioritization of genomes to be sequenced. This prioritization...
Genome-Wide Requirements for Resistance to Functionally Distinct DNA-Damaging Agents
Lee, William, St.Onge, Robert P, Proctor, Michael, Flaherty, Patrick, Jordan, Michael I, Arkin, Adam P, ...
The mechanistic and therapeutic differences in the cellular response to DNA-damaging compounds are not completely understood, despite intense study. To expand our knowledge of DNA damage, we assayed...
Protein Molecular Function Prediction by Bayesian Phylogenomics
Engelhardt, Barbara E, Jordan, Michael I, Muratore, Kathryn E, Brenner, Steven E
We present a statistical graphical model to infer specific molecular function for unannotated protein sequences using homology. Based on phylogenomic principles, SIFTER (Statistical Inference of...
Hierarchical Dirichlet Processes
Teh, Yee Whye, Jordan, Michael I., Beal, Matthew J., Blei, David M.
A critical assessment of Mus musculus gene function prediction using integrated genomic evidence
Peña-Castillo, Lourdes, Tasan, Murat, Myers, Chad L, Lee, Hyunju, Joshi, Trupti, Zhang, Chao, ...
Consistent probabilistic outputs for protein function prediction
Obozinski, Guillaume, Lanckriet, Gert, Grant, Charles, Jordan, Michael I, Noble, William Stafford
In predicting hierarchical protein function annotations, such as terms in the Gene Ontology (GO), the simplest approach makes predictions for each term independently. However, this approach has the...
A Dual Receptor Crosstalk Model of G-Protein-Coupled Signal Transduction
Flaherty, Patrick, Radhakrishnan, Mala L., Dinh, Tuan, Rebres, Robert A., Roach, Tamara I., Jordan, Michael I., ...
Macrophage cells that are stimulated by two different ligands that bind to G-protein-coupled receptors (GPCRs) usually respond as if the stimulus effects are additive, but for a minority of ligand...
On the inference of ancestries in admixed populations
Sankararaman, Sriram, Kimmel, Gad, Halperin, Eran, Jordan, Michael I.
Inference of ancestral information in recently admixed populations, in which every individual is composed of a mixed ancestry (e.g., African Americans in the United States), is a challenging problem....
Attractor Dynamics in Feedforward Neural Networks
Lawrence Saul Michael, Michael I. Jordan
We study the probabilistic generative models parameterized by feedforward neural networks. An attractor dynamics for probabilistic inference in these models is derived from a mean field approximation...
Attractor Dynamics in Feedforward Neural Networks
Lawrence Saul Michael, Michael I. Jordan
We study the probabilistic generative models parameterized by feedforward neural networks. An attractor dynamics for probabilistic inference in these models is derived from a mean field approximation...
Modular and Hierarchical Learning Systems
Ma Mit Press, Robert A. Jacobs, Michael I. Jordan, Michael I. Jordan
this article we discuss the problem of learning in modular and hierarchical systems. Modular and hierarchical systems allow complex learning problems to be solved by dividing the problem into a set...
Attractor Dynamics in Feedforward Neural Networks
Lawrence Saul, Michael I. Jordan
We study the probabilistic generative models parameterized by feedforward neural networks. An attractor dynamics for probabilistic inference in these models is derived from a mean field approximation...
Recursive Algorithms for Approximating Probabilities in Graphical Models
Tommi Jaakkola, Michael I. Jordan
We develop a recursive node-elimination formalism for efficiently approximating large probabilistic networks. No constraints are set on the network topologies. Yet the formalism can be...
Attractor Dynamics in Feedforward Neural Networks
Lawrence K. Saul, Michael I. Jordan
We study the probabilistic generative models parameterized by feedforward neural networks. An attractor dynamics for probabilistic inference in these models is derived from a mean field approximation...
Convexity, Classification, and Risk Bounds
Peter Bartlett Bartlett, Michael I. Jordan, Jon D. Mcauliffe
Many of the classification algorithms developed in the machine learning literature, including the support vector machine and boosting, can be viewed as minimum contrast methods that minimize a convex...
Convexity, Classification, and Risk Bounds
Peter Bartlett Bartlett, Michael I. Jordan, Jon D. Mcauliffe
Many of the classification algorithms developed in the machine learning literature, including the support vector machine and boosting, can be viewed as minimum contrast methods that minimize a convex...
Association Mapping and Significance Estimation via the Coalescent
Kimmel, Gad, Karp, Richard M., Jordan, Michael I., Halperin, Eran
The central questions asked in whole-genome association studies are how to locate associated regions in the genome and how to estimate the significance of these findings. Researchers usually do this...
Joint estimation of gene conversion rates and mean conversion tract lengths from population SNP data
Yin, Junming, Jordan, Michael I., Song, Yun S.
Motivation: Two known types of meiotic recombination are crossovers and gene conversions. Although they leave behind different footprints in the genome, it is a challenging task to tease apart their...