Optimization Techniques for Semi-Supervised Support Vector Machines (2008)
Olivier Chapelle, Vikas Sindhwani, Sathiya S. Keerthi, Nello Cristianini
Due to its wide applicability, the problem of semi-supervised classification is attracting increasing attention in machine learning. Semi-Supervised Support Vector Machines (S 3 VMs) are based on...
A Fast Tracking Algorithm for Generalized LARS/LASSO (2007)
Keerthi, Sathiya S, Shevade, Shirish
This letter gives an efficient algorithm for tracking the solution curve of sparse logistic regression with respect to the regularization parameter. The algorithm is based on approximating the...
Fast Generalized Cross-Validation Algorithm for Sparse Model Learning (2007)
Sundararajan, S, Shevade, Shirish, Keerthi, Sathiya S
We propose a fast, incremental algorithm for designing linear regression models. The proposed algorithm generates a sparse model by optimizing multiple smoothing parameters using the generalized...
Fast Generalized Cross-Validation Algorithm for Sparse Model Learning (2007)
Sundararajan, S, Shevade, Shirish, Keerthi, Sathiya S
We propose a fast, incremental algorithm for designing linear regression models. The proposed algorithm generates a sparse model by optimizing multiple smoothing parameters using the generalized...
Multi-category classification by soft-max combination of binary classifiers (2003)
Duan, Kaibo, Keerthi, Sathiya S, Chu, Wei, Shevade, Shirish Krishnaj, Poo, Aun Neow
In this paper, we propose a multi-category classification method that combines binary classifiers through soft-max function. Posteriori probabilities are also obtained. Both, one-versus-all and...
Babu, Phanendra G, Murty, Narasimha M, Keerthi, Sathiya S
In this paper, a stochastic connectionist approach is proposed for solving function optimization problems with real-valued parameters. With the assumption of increased processing capability of a node...
Information geometry and Plefka’s mean-field theory (2000)
Bhattacharyya, C, Keerthi, Sathiya S
An alternate derivation of Plefka’s method is presented which is obtained from minimizing Gibbs energy. It is shown that this method can be rederived using a perturbation expansion of...
Babu, Phanendra G, Murty, Narasimha M, Keerthi, Sathiya S
In this paper, a stochastic connectionist approach is proposed for solving function optimization problems with real-valued parameters. With the assumption of increased processing capability of a node...
Information geometry and Plefka’s mean-field theory (2000)
Bhattacharyya, C, Keerthi, Sathiya S
An alternate derivation of Plefka’s method is presented which is obtained from minimizing Gibbs energy. It is shown that this method can be rederived using a perturbation expansion of...
Synthesis of Fault-Tolerant Feedforward Neural Networks Using Minimax Optimization (1998)
Deodhare, Dipti, Vidyasagar, M, Keerthi, Sathiya S
In this paper we examine a technique by which fault tolerance can be embedded into a feedforward network leading to a network tolerant to the loss of a node and its associated weights. The fault...
Synthesis of Fault-Tolerant Feedforward Neural Networks Using Minimax Optimization (1998)
Deodhare, Dipti, Vidyasagar, M, Keerthi, Sathiya S
In this paper we examine a technique by which fault tolerance can be embedded into a feedforward network leading to a network tolerant to the loss of a node and its associated weights. The fault...
Gain optimization under control structure and stability region constraints (1996)
Phatak, Makarand S, Keerthi, Sathiya S
In this paper we consider the problem of optimizing the Frobenious norm of linear static feedback gains of a system with control structure and stability region constraints. We consider four important...
Gain optimization under control structure and stability region constraints (1996)
Phatak, Makarand S, Keerthi, Sathiya S
In this paper we consider the problem of optimizing the Frobenious norm of linear static feedback gains of a system with control structure and stability region constraints. We consider four important...
Dattasharma, Abhi, Keerthi, Sathiya S
This paper concerns the development of a piecewise linear Voronoi roadmap for translating a convex polyhedron in a three-dimensional (3-D) polyhedral world. In general the Voronoi roadmap is...
Regional Pole Placement of Multivariable Systems Under Control Structure Constraints (1995)
Keerthi, Sathiya S, Phatak, Makarand S
Many controller realizations are structurally constrained. Some typical examples are static output feedback, constant gain feedback for multiple operating points of a system, two-controller feedback,...
Regional Pole Placement of Multivariable Systems Under Control Structure Constraints (1995)
Keerthi, Sathiya S, Phatak, Makarand S
Many controller realizations are structurally constrained. Some typical examples are static output feedback, constant gain feedback for multiple operating points of a system, two-controller feedback,...
Dattasharma, Abhi, Keerthi, Sathiya S
This paper concerns the development of a piecewise linear Voronoi roadmap for translating a convex polyhedron in a three-dimensional (3-D) polyhedral world. In general the Voronoi roadmap is...
Numerical Determination of Optimal Non-Holonomic Paths in the Presence of Obstacles (1993)
Samuel, Sudhaker, Keerthi, Sathiya S
This paper addresses the problem of numerically finding an optimal path for a robot with non-holonomic constraints. A car like robot, whose turning radius is lower bounded is considered as an...
A Nonlinear Least Squares Approach to the Numerical Optimal Control of Non-holonomic Systems (1993)
Samuel, Sudhaker, Keerthi, Sathiya S
This paper gives a nonlinear least squares approach for numerically finding a trajectory to transfer a non-holonomic system from one configuration to another while satisfying given point-wise...
Numerical Determination of Optimal Non-Holonomic Paths in the Presence of Obstacles (1993)
Samuel, Sudhaker, Keerthi, Sathiya S
This paper addresses the problem of numerically finding an optimal path for a robot with non-holonomic constraints. A car like robot, whose turning radius is lower bounded is considered as an...
A Nonlinear Least Squares Approach to the Numerical Optimal Control of Non-holonomic Systems (1993)
Samuel, Sudhaker, Keerthi, Sathiya S
This paper gives a nonlinear least squares approach for numerically finding a trajectory to transfer a non-holonomic system from one configuration to another while satisfying given point-wise...
Phatak, Makarand S, Keerthi, Sathiya S
Various structural Constraints on the state feedback gain matrix, K, used in the control of a multivariable dynamical system arise because of the way feedback is realized. Static output feedback,...
Venugopal, K, Keerthi, Sathiya S
The problem of numerically computing an approximate parameterization of the curve of intersection of (n-1) surfaces in $R^n$ is considered. A precise definition of an ϵ-approximate...
Phatak, Makarand S, Keerthi, Sathiya S
Various structural Constraints on the state feedback gain matrix, K, used in the control of a multivariable dynamical system arise because of the way feedback is realized. Static output feedback,...
Venugopal, K, Keerthi, Sathiya S
The problem of numerically computing an approximate parameterization of the curve of intersection of (n-1) surfaces in $R^n$ is considered. A precise definition of an ϵ-approximate...
This thesis deals with the optimal regulation of constrained discrete-time systems. The class of problems considered is characterized by state-control constraints, a general nonlinear cost function...