The STUDIA UNIVERSITATIS BABEŞ-BOLYAI issue article summary

The summary of the selected article appears at the bottom of the page. In order to get back to the contents of the issue this article belongs to you have to access the link from the title. In order to see all the articles of the archive which have as author/co-author one of the authors mentioned below, you have to access the link from the author's name.

  
       
         
    STUDIA INFORMATICA - Issue no. Sp.Issue%201 / 2009  
         
  Article:   Q-LEARNING AND POLICY GRADIENT METHODS.

Authors:  HUNOR JAKAB, LEHEL CSATÓ. 		 
       
         
  Abstract:  Many real-world tasks require a robotic agent to adapt its behavior to certain environmental conditions and to acquire knowledge without user interaction. In reinforcement learning knowledge is usually acquired without preexisting training data, there by making the learning process more "natural". In this paper we investigate two reinforcement learning methods and present a simulation environment where we test their performance. The simulation environment allows the testing of various reinforcement algorithms without a need for the physical robot. Its advantage is that it can be used to perform benchmarks and evaluations of different learning algorithms.

Key words and phrases. reinforcement learning, Markov decision processes, robotics.

  
         
     
         
         
      Back to previous page