Combining learning algorithms: An approach to Markov decision processes

Combining learning algorithms: An approach to Markov decision processes

Ribeiro, Richardson and Favarim, Fábio and Barbosa, Marco A.C. and Koerich, Alessandro L. and Enembreck, Fabrício

Lecture Notes in Business Information Processing 2013

Abstract : In this paper we present a technique for estimating policies which combines instance-based learning and reinforcement learning algorithms in Markovian environments. This approach has been developed for speeding up the convergence of adaptive intelligent agents that using reinforcement learning algorithms. Speeding up the learning of an intelligent agent is a complex task since the choice of inadequate updating techniques may cause delays in the learning process or even induce an unexpected acceleration that causes the agent to converge to a non-satisfactory policy. Experimental results in real-world scenarios have shown that the proposed technique is able to speed up the convergence of the agents while achieving optimal policies, overcoming problems of classical reinforcement learning approaches.