Improving Gradient Estimation by Incorporating Sensor Data (2009)
Gregory Lawrence, Stuart Russell
An efficient policy search algorithm should estimate the local gradient of the objective function, with respect to the policy parameters, from as few trials as possible. Whereas most policy search...
1 BLOG: Probabilistic Models with Unknown Objects (2008)
Brian Milch, Bhaskara Marthi, Stuart Russell, David Sontag, Daniel L. Ong, Andrey Kolobov
Human beings and AI systems must convert sensory input into some understanding of what is going on in the world around them. That is, they must make inferences about the objects and events that...
Abstract Identity Uncertainty (2008)
We are often uncertain about the identity of objects. This phenomenon appears in theories of object persistence in early childhood; in the well-known Morning Star/Evening Star example; in tracking...
Optimal composition of real-time systems (2008)
Shlomo Zilbersteinyand, Stuart Russell
Real-time systems are designed for environments in which the utility of actions is strongly time-dependent. Recent work by Dean, Horvitz and others has shown that anytime algorithms are a useful tool...
Jeffrey Forbes, Nikunj Oza, Ronald Parr, Stuart Russell
This project investigated the feasibility of constructing an autonomous vehicle controller based on probabilistic inference and utility maximization. We believed, and still believe, that such methods...
Probabilistic Modeling of Sensor Artifacts in Critical Care (2008)
Manley, Geoffrey, Staudenmayer, Kristan, Cohen, Mitchell, Madden, Michael G., Morabito, Diane, Aleks, Norm, ...
We describe an application of probabilistic modeling and inference technology to the problem of analyzing sensor data in the setting of an intensive care unit (ICU). In particular, we consider the...
Probabilistic Detection of Short Events, with Application to Critical Care Monitoring (2008)
Manley, Geoffrey, Cohen, Mitchell, Staudenmayer, Kristan, Morabito, Diane, Madden, Michael G., Russell, Stuart, ...
We describe an application of probabilistic modeling and inference technology to the problem of analyzing sensor data in the setting of an intensive care unit (ICU). In particular, we consider the...
Handbook of Perception and Cognition, Vol.14 Chapter 4: Machine Learning (2007)
Stuart Russell, I Introduction
Contents I Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 A A general model of learning . . . . . . . . . . . . . . . . . . . . . . ....
By Isaac, Isaac Cheng, Taught Professor, Stuart Russell
.................................................................................................................1 Scheduling...
Reviewed By, Stuart Russell, Eric Welfald, Simon Parsons, London Wca Px
rse, in any normal search there is some such deliberation when algorithms such as A* are used to try to ensure that only nodes that are liable to lead most cheaply to a goal are expanded. However,...
Technical Report ??? UC Berkeley VARIATIONAL MCMC (2007)
Nando De Freitas, Stuart Russell
We propose a new class of learning algorithms that combines variational approximation and Markov chain Monte Carlo (MCMC) simulation. One of these algorithms is a mixture of two MCMC kernels: a...
S.: First-order probabilistic languages: Into the unknown (2007)
Abstract. This paper surveys first-order probabilistic languages (FOPLs), which combine the expressive power of first-order logic with a probabilistic treatment of uncertainty. We provide a taxonomy...
S.: First-order probabilistic languages: Into the unknown (2007)
Abstract. This paper surveys first-order probabilistic languages (FOPLs), which combine the expressive power of first-order logic with a probabilistic treatment of uncertainty. We provide a taxonomy...
Improving Gradient Estimation by Incorporating Sensor Data (2006)
Gregory Donnell Lawrence, Stuart J. Russell, Gregory Lawrence, Stuart Russell
personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the...
BLOG: Probabilistic Models with Unknown Objects (2006)
Milch, Brian, Marthi, Bhaskara, Russell, Stuart, Sontag, David, Ong, Daniel L., Kolobov, Andrey
We introduce BLOG, a formal language for defining probability models with unknown objects and identity uncertainty. A BLOG model describes a generative process in which some steps add objects to the...
Approximate Inference for Infinite Contingent Bayesian Networks (2005)
Brian Milch, Bhaskara Marthi, David Sontag, Stuart Russell, Daniel L. Ong, Andrey Kolobov
In many practical problems---from tracking aircraft based on radar data to building a bibliographic database based on citation lists---we want to reason about an unbounded number of unseen objects...
Concurrent Hierarchical Reinforcement Learning (2005)
Bhaskara Marthi Stuart, Stuart Russell, David Latham
We consider applying hierarchical reinforcement learning techniques to problems in which an agent has several effectors to control simultaneously. We argue that the kind of prior knowledge one...
Blog: Probabilistic models with unknown objects (2005)
Brian Milch, Bhaskara Marthi, Stuart Russell, David Sontag, Daniel L. Ong, Andrey Kolobov
This paper introduces and illustrates BLOG, a formal language for defining probability models over worlds with unknown objects and identity uncertainty. BLOG unifies and extends several existing...
Markov Chain Monte Carlo Data Association for General Multiple-Target Tracking Problems (2004)
Songhwai Oh, Stuart Russell, Shankar Sastry
In this paper, we consider the general multipletarget tracking problem in which an unknown number of targets appears and disappears at random times and the goal is to find the tracks of targets from...
Markov Chain Monte Carlo Data Association for General Multiple-Target Tracking Problems (2004)
Songhwai Oh, Stuart Russell, Shankar Sastry
In this paper, we consider the general multipletarget tracking problem in which an unknown number of targets appears and disappears at random times and the goal is to find the tracks of targets from...
Blog: Relational modeling with unknown objects (2004)
Brian Milch, Bhaskara Marthi, Stuart Russell
In many real-world probabilistic reasoning problems, one of the questions we want to answer is: how many objects are out there? Examples of such problems range from multitarget tracking to extracting...
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...
Eyal Amir And, Eyal Amir, Stuart Russell
Filtering denotes any method whereby an agent updates its belief state---its knowledge of the state of the world---from a sequence of actions and observations.
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...
Filtering denotes any method whereby an agent updates its belief state—its knowledge of the state of the world—from a sequence of actions and observations. In logical filtering, the belief state...
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...
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...
Identity uncertainty and citation matching (2003)
Hanna Pasula, Bhaskara Marthi, Brian Milch, Stuart Russell, Ilya Shpitser
Identity uncertainty is a pervasive problem in real-world data analysis. It arises whenever objects are not labeled with unique identifiers or when those identifiers may not be perceived perfectly....
First-Order Probabilistic Models for Information Extraction (2003)
Bhaskara Marthi, Brian Milch, Stuart Russell
Information extraction (IE) is the problem of constructing a knowledge base from a corpus of text documents. In this paper, we argue that firstorder probabilistic models (FOPMs) are a promising...
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...
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...
Ecient Gradient Estimation for Motor Control Learning (2003)
Gregory Lawrence, Noah Cowan, Stuart Russell
The task of estimating the gradient of a function in the presence of noise is central to several forms of reinforcement learning, including policy search methods. We present two techniques for...
Filtering denotes any method whereby an agent updates its belief state—its knowledge of the state of the world—from a sequence of actions and observations. In logical filtering, the belief state...
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...
Statistical visual language models for ink parsing (2002)
Michael Shilman, Hanna Pasula, Stuart Russell, Richard Newton
Abstract 1 In this paper we motivate a new technique for automatic recognition of hand-sketched digital ink. By viewing sketched drawings as utterances in a visual language, sketch recognition can be...
Programmable reinforcement learning agents (2001)
Stuart Russell, Andrew L. Zimdars
The paper explores a very simple agent design method called Q-decomposition, wherein a complex agent is built from simpler subagents. Each subagent has its own reward function and runs its own...
Online bagging and boosting (2001)
Abstract Bagging and boosting are well-known ensemble learning methods. They combine multiple learned base models with the aim of improving generalization performance. To date, they have been used...
Rao-Blackwellised Particle Filtering for Dynamic Bayesian Networks (2000)
Arnaud Doucet, Nando De Freitas, Kevin Murphy, Stuart Russell
Particle filters (PFs) are powerful sampling-based inference/learning algorithms for dynamic Bayesian networks (DBNs). They allow us to treat, in a principled way, any type of probability...
Algorithms for Inverse Reinforcement Learning (2000)
This paper addresses the problem of inverse reinforcement learning (IRL) in Markov decision processes, that is, the problem of extracting a reward function given observed, optimal behaviour. IRL may...
Algorithms for Inverse Reinforcement Learning (2000)
This paper addresses the problem of inverse reinforcement learning (IRL) in Markov decision processes, that is, the problem of extracting a reward function given observed, optimal behavior. IRL may...
Rao-Blackwellised Particle Filtering for Dynamic Bayesian Networks (2000)
Arnaud Doucet Engineering, Arnaud Doucet, Engineering Dept, Nando De Freitas, Kevin Murphy, Stuart Russell
Particle filters (PFs) are powerful samplingbased inference/learning algorithms for dynamic Bayesian networks (DBNs). They allow us to treat, in a principled way, any type of probability...
Tracking Many Objects With Many Sensors (1999)
Hanna Pasula Stuart, Stuart Russell, Michael Ostl
Keeping track of multiple objects over time is a problem that arises in many real-world domains. The problem is often complicated by noisy sensors and unpredictable dynamics. Previous work by Huang...
Policy invariance under reward transformations: Theory and application to reward shaping (1999)
Andrew Y. Ng, Daishi Harada, Stuart Russell
This paper investigates conditions under which modifications to the reward function of a Markov decision process preserve the optimal policy. It is shown that, besides the positive linear...
Learning the Structure of Dynamic Probabilistic Networks (1998)
Nir Friedman, Kevin Murphy, Stuart Russell
Dynamic probabilistic networks are a compact representation of complex stochastic processes. In this paper we examine how to learn the structure of a DPN from data. We extend structure scoring rules...
Probabilistic Modeling With Bayesian Networks For Asr (1998)
Geoffrey Zweig, Stuart Russell
This paper describes the application of Bayesian networks to automatic speech recognition. Bayesian networks enable the construction of probabilistic models in which an arbitrary set of variables can...
Richard Dearden, Nir Friedman, Stuart Russell
A central problem in learning in complex environments is balancing exploration of untested actions against exploitation of actions that are known to be good. The benefit of exploration can be...
Learning agents for uncertain environments (Extended Abstract) (1998)
This talk proposes a very simple "baseline architecture " for a learning agent that can handle stochastic, partially observable environments. The architecture uses reinforcement learning...
Object Identification: A Bayesian Analysis with Application to Traffic Surveillance (1998)
Object identification---the task of deciding that two observed objects are in fact one and the same object---is a fundamental requirement for any situated agent that reasons about individuals. Object...
Speech Recognition with Dynamic Bayesian Networks (1998)
Geoffrey Zweig, Stuart Russell
Dynamic Bayesian networks (DBNs) are a useful tool for representing complex stochastic processes. Recent developments in inference and learning in DBNs allow their use in real-world applications. In...
this document is to derive the algorithm in its most general form from first principles and to give a short proof of its convergence. The derivation extends the mixture-model derivation from Bishop...
Richard Dearden, Nir Friedman, Stuart Russell
A central problem in learning in complex environments is balancing exploration of untested actions against exploitation of actions that are known to be good. The benefit of exploration can be...
Learning the structure of dynamic probabilistic networks (1998)
Nir Friedman, Kevin Murphy, Stuart Russell
Dynamic probabilistic networks are a compact representation of complex stochastic processes. In this paper we examine how to learn the structure of a DPN from data. We extend structure scoring rules...
Image segmentation in video sequences: A probabilistic approach (1997)
"Background subtraction" is an old technique for finding moving objects in a video sequence---for example, cars driving on a freeway. The idea is that subtracting the current image from a...
Compositional Modeling with DPNs (1997)
Geoffrey Zweig, Geoffrey Zweig, Stuart Russell, Stuart Russell
Dynamic probabilistic networks (DPNs) are a powerful and efficient method for encoding stochastic temporal models. In the past, however, their use has been largely confined to the description of...
Object Identification in a Bayesian Context (1997)
Object identification---the task of deciding that two observed objects are in fact one and the same object---is a fundamental requirement for any situated agent that reasons about individuals. Object...
Adaptive Probabilistic Networks with Hidden Variables (1997)
John Binder, Daphne Koller, Stuart Russell, Keiji Kanazawa, Padhraic Smyth
. Probabilistic networks (also known as Bayesian belief networks) allow a compact description of complex stochastic relationships among several random variables. They are rapidly becoming the tool of...
Adaptive Probabilistic Networks with Hidden Variables (1997)
John Binder, Daphne Koller, Stuart Russell, Keiji Kanazawa, Padhraic Smyth
. Probabilistic networks (also known as Bayesian belief networks) allow a compact description of complex stochastic relationships among several random variables. They are used widely for uncertain...
Reinforcement Learning with Hierarchies of Machines (1997)
We present a new approach to reinforcement learning in which the policies considered by the learning process are constrained by hierarchies of partially specified machines. This allows for the use of...
Rationality and intelligence (1997)
The long-term goal of our field is the creation and understanding of intelligence. Productive research in AI, both practical and theoretical, benefits from a notion of intelligence that is precise...
Jitendra Malik, Stuart Russell
this report. Computing reliable matches is the central task and presents a number of difficult technical problems. The approach described below achieves good performance even under difficult...
Optimal Composition of Real-Time Systems (1996)
Shlomo Zilberstein, Stuart Russell
Real-time systems are designed for environments in which the utility of actions is strongly time-dependent. Recent work by Dean, Horvitz and others has shown that anytime algorithms are a useful tool...
Local Learning in Probabilistic Networks With Hidden Variables (1995)
Stuart Russell, Keiji Kanazawa
Probabilistic networks, which provide compact descriptions of complex stochastic relationships among several random variables, are rapidly becoming the tool of choice for uncertain reasoning in...
Rationality and Intelligence (1995)
The long-term goal of our field is the creation and understanding of intelligence. Productive research in AI, both practical and theoretical, benefits from a notion of intelligence that is precise...
The BATmobile: Towards a Bayesian Automated Taxi (1995)
Jeff Forbes, Tim Huang, Keiji Kanazawa, Stuart Russell
The problem of driving an autonomous vehicle in normal traffic engages many areas of AI research and has substantial economic significance. We describe work in progress on a new approach to this...
Provably Bounded-Optimal Agents (1995)
Stuart Russell, Devika Subramanian
Since its inception, artificial intelligence has relied upon a theoretical foundation centred around perfect rationality as the desired property of intelligent systems. We argue, as others have done,...
Approximating Optimal Policies for Partially Observable Stochastic Domains (1995)
The problem of making optimal decisions in uncertain conditions is central to Artificial Intelligence. If the state of the world is known at all times, the world can be modeled as a Markov Decision...
Rationality and Intelligence (1995)
The long-term goal of our field is the creation and understanding of intelligence. Productive research in AI, both practical and theoretical, benefits from a notion of intelligence that is precise...
Approximating Optimal Policies for Partially Observable Stochastic Domains (1995)
The problem of making optimal decisions in uncertain conditions is central to Artificial Intelligence. If the state of the world is known at all times, the world can be modeled as a Markov Decision...
Probabilistic models based on directed acyclic graphs (DAGs) have a long and rich tradition, which began with the geneticist Sewall Wright (1921). Variants have apeared in many elds � within...
Local learning in probabilistic networks with hidden variables (1995)
Stuart Russell, John Binder, Daphne Koller, Keiji Kanazawa
Probabilistic networks which provide compact descriptions of complex stochastic relationships among several random variables are rapidly becoming the tool of choice for uncertain reasoning in...
The batmobile: Towards a bayesian automated taxi (1995)
Jeff Forbes, Tim Huang, Keiji Kanazawa, Stuart Russell
The problem of driving an autonomous vehicle in normal traffic engages many areas of AI research and has substantial economic significance. We describe work in progress on a new approach to this...
The batmobile: Towards a bayesian automated taxi (1995)
Jeff Forbes, Tim Huang, Keiji Kanazawa, Stuart Russell
The problem of driving an autonomous vehicle in normal traffic engages many areas of AI research and has substantial economic significance. We describe work in progress on a new approach to this...
Control Strategies for a Stochastic Planner (1994)
We present new algorithms for local planning over Markov decision processes. The base-level algorithm possesses several interesting features for control of computation, based on selecting...
Adaptive Probabilistic Networks (1994)
Stuart Russell, John Binder, Daphne Koller
Belief networks (or probabilistic networks) and neural networks are two forms of network representations that have been used in the development of intelligent systems in the field of artificial...
NP-Completeness of Searches for Smallest Possible Feature Sets (1994)
In many learning problems, the learning system is presented with values for features that are actually irrelevant to the concept it is trying to learn. The FOCUS algorithm, due to Almuallim and...
Judea Pearl Computer, Stuart Russell, J. Pearl, S. Russell
INTRODUCTION Probabilistic models based on directed acyclic graphs have a long and rich tradition, beginning with work by the geneticist Sewall Wright in the 1920s. Variants have appeared in many...
Techniques for handling inference complexity in Dynamic Belief Networks (1993)
Ann Nicholson, Ann Nicholson, Stuart Russell, Stuart Russell
Dynamic Belief Networks (DBNs) have been of interest recently as modelling tools for environments that change over time. In such networks, sets of nodes may be added automatically over time to...
Efficient Memory-Bounded Search Methods (1992)
. Memory-bounded algorithms such as Korf's IDA* and Chakrabarti et al's MA* are designed to overcome the impractical memory requirements of heuristic search algorithms such as A* . It is...
PAC-Learnability of Determinate Logic Programs (1992)
Saso Dzeroski, Stephen Muggleton, Stuart Russell
The field of Inductive Logic Programming (ILP) is concerned with inducing logic programs from examples in the presence of background knowledge. This paper defines the ILP problem, and describes the...
Principles of Metareasoning (1991)
Stuart Russell, Eric Wefald, Maurice Karnaugh, Richard Karp, David Mcallester, Devika Subramanian, ...
In this paper we outline a general approach to the study of metareasoning, not in the sense of explicating the semantics of explicitly specified meta-level control policies, but in the sense of...
Space-Efficient Inference in Dynamic Probabilistic Networks
John Binder, Kevin Murphy, Stuart Russell
Dynamic probabilistic networks (DPNs) are a useful tool for modeling complex stochastic processes. The simplest inference task in DPNs is monitoring --- that is, computing a posterior distribution...