A Tutorial for Reinforcement Learning

Description

A Tutorial for Reinforcement LearningAbhijit GosaviDepartment of Engineering Management and Systems EngineeringMissouri University of Science and Technology219 Engineering Management, Rolla, MO 65409Email:[e-mail] 16, 20091Contents1 Introduction 32 MDPs and SMDPs 33 RL 63.1 Average reward . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73.2 Selecting the learning rate or step size . . . . . . . . . . . . . . . . . . . . . 93.3 Discounted reward . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103.4 Codes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 Conclusions 1121 IntroductionThe tutorial is written for those who would like an introduction to reinforcement learning(RL). The aim is to provide an intuitive presentation of the ideas rather than concentrateon the deeper mathematics.RL is generally used to solve the so-called Markov decision problem (MDP). In otherwords, the problem that you are attempting to solve with RL should be an MDP or itsvariant. The theory of RL relies on dynamic programming (DP) and artiflcial intelligence(AI). We will begin with a quick description of MDPs. We will discuss what we mean by\complex"and\large-scale"MDPs. ThenwewillexplainwhyRLisneededtosolvecomplexand large-scale MDPs. The semi-Markov decision problem (SMDP) will also be covered.The tutorial is meant to serve as an introduction to these topics and is based mostly onthe ...

Informations

• Publié par: Edwie
• Nombre de lectures: 17
• Langue: English

Extrait

A Tutorial for Reinforcement Learning

Abhijit Gosavi Department of Engineering Management and Systems Engineering Missouri University of Science and Technology 219 Engineering Management, Rolla, MO 65409 Email:[e-mail]

December 16, 2009

1