Modeling reinforcement learning algorithms for performance analysis

Authors:
Shrirang Ambaji Kulkarni;G Raghavendra Rao
Affiliations:
Gogte Institute of Technology, Belgaum, India;National Institute of Engineering, Mysore, India
Venue:
Proceedings of the International Conference on Advances in Computing, Communication and Control
Year:
2009

Citing 6
Cited 0

Technical Note: \cal Q-Learning

Machine Learning
Efficient learning and planning within the Dyna framework

Proceedings of the second international conference on From animals to animats 2 : simulation of adaptive behavior: simulation of adaptive behavior
Prioritized Sweeping: Reinforcement Learning with Less Data and Less Time

Machine Learning
Recent Advances in Reinforcement Learning

Recent Advances in Reinforcement Learning
Introduction to Reinforcement Learning

Introduction to Reinforcement Learning
Artificial Intelligence: Structures and Strategies for Complex Problem Solving (5th Edition)

Artificial Intelligence: Structures and Strategies for Complex Problem Solving (5th Edition)

Quantified Score

Hi-index	0.00

Visualization

Abstract

Reinforcement Learning Algorithms present interesting learning techniques. Here an autonomous agent interacts with its environment to choose optimal actions to achieve its goals. The performance of an agent is determined by how quickly it learns and converges to an optimal solution. Q-learning and Prioritized sweeping provide interesting techniques to achieve this. In this paper we try to analyze the performance of Q-learning and Prioritized sweeping as examples of model free and model based reinforcement learning. We also try to analyze the optimal number of backups required for prioritized sweeping. We model the results of prioritized sweeping as a regression model and discuss the prediction of the model by comparing it with the accuracy of our simulation results.