Convergence Results for Single-Step On-Policy Reinforcement-Learning Algorithms

• 1 January 2000
• journal article
• Published by Springer Nature in Machine Learning
• Vol. 38  (3) , 287-308
• https://doi.org/10.1023/a:1007678930559

Abstract

Keywords

• CONVERGENCE RESULTS
• EXAMPLES OFEXPLORATION STRATEGY
• OPTIMAL VALUE
• SINGLE-STEP ON-POLICYREINFORCEMENT-LEARNING ALGORITHMS
• EXPLORATION PROBLEM
• RELATED ON-POLICY ALGORITHM
• EXPLOITATION TRADEOFF
• SEPARATE EXPLORATION
• PERSISTENT EXPLORATION
• CONTROL PROBLEM
• REINFORCEMENT LEARNING
• OPTIMAL POLICY
• MARKOV DECISION PROCESSES
• MARKOV DECISION PROCESS
• CONVERGENCE