Sciweavers

ICML
2003
IEEE
15 years 15 days ago
TD(0) Converges Provably Faster than the Residual Gradient Algorithm
In Reinforcement Learning (RL) there has been some experimental evidence that the residual gradient algorithm converges slower than the TD(0) algorithm. In this paper, we use the ...
Ralf Schoknecht, Artur Merke
ICML
2003
IEEE
15 years 15 days ago
Optimization with EM and Expectation-Conjugate-Gradient
We show a close relationship between the Expectation - Maximization (EM) algorithm and direct optimization algorithms such as gradientbased methods for parameter learning. We iden...
Ruslan Salakhutdinov, Sam T. Roweis, Zoubin Ghahra...
ICML
2003
IEEE
15 years 15 days ago
Relativized Options: Choosing the Right Transformation
Relativized options combine model minimization methods and a hierarchical reinforcement learning framework to derive compact reduced representations of a related family of tasks. ...
Balaraman Ravindran, Andrew G. Barto
ICML
2003
IEEE
15 years 15 days ago
Adaptive Overrelaxed Bound Optimization Methods
We study a class of overrelaxed bound optimization algorithms, and their relationship to standard bound optimizers, such as ExpectationMaximization, Iterative Scaling, CCCP and No...
Ruslan Salakhutdinov, Sam T. Roweis
ICML
2003
IEEE
15 years 15 days ago
Adaptive Feature-Space Conformal Transformation for Imbalanced-Data Learning
When the training instances of the target class are heavily outnumbered by non-target training instances, SVMs can be ineffective in determining the class boundary. To remedy this...
Gang Wu, Edward Y. Chang
ICML
2003
IEEE
15 years 15 days ago
Tackling the Poor Assumptions of Naive Bayes Text Classifiers
Naive Bayes is often used as a baseline in text classification because it is fast and easy to implement. Its severe assumptions make such efficiency possible but also adversely af...
Jason D. Rennie, Lawrence Shih, Jaime Teevan, Davi...
ICML
2003
IEEE
15 years 15 days ago
Q-Decomposition for Reinforcement Learning Agents
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...
Stuart J. Russell, Andrew Zimdars
ICML
2003
IEEE
15 years 15 days ago
Stochastic Local Search in k-Term DNF Learning
A novel native stochastic local search algorithm for solving k-term DNF problems is presented. It is evaluated on hard k-term DNF problems that lie on the phase transition and com...
Stefan Kramer, Ulrich Rückert
ICML
2003
IEEE
15 years 15 days ago
Kernel PLS-SVC for Linear and Nonlinear Classification
A new method for classification is proposed. This is based on kernel orthonormalized partial least squares (PLS) dimensionality reduction of the original data space followed by a ...
Roman Rosipal, Leonard J. Trejo, Bryan Matthews
ICML
2003
IEEE
15 years 15 days ago
Weighted Order Statistic Classifiers with Large Rank-Order Margin
We investigate how stack filter function classes like weighted order statistics can be applied to classification problems. This leads to a new design criteria for linear classifie...
Reid B. Porter, Damian Eads, Don R. Hush, James Th...