ABSTRACT
In this work, we present a reinforcement-based learning algorithm that includes the automatic
classification of both sensors and actions. The classification process is prior to any application of
reinforcement learning. If categories are not at the adequate abstraction level, the problem could be not
learnable. The classification process is usually done by the programmer and is not considered as part
of the learning process. However, in complex tasks, environments, or agents, this manual process
could become extremely difficult. To solve this inconvenience, we propose to include the
classification into the learning process. We apply an algorithm to automatically learn to achieve a task
through reinforcement learning that works without needing a previous classification process. The
system is called Fish or Ship (FOS) assigned the task of inducing classification rules for classification
task described in terms of 6 attributes. The task is to categorize an object that has one or more of the
following features: Sail, Solid, Big, Swim, Eye, Fins into one of the following: fish, or ship. First
results of the application of this algorithm are shown Reinforcement learning techniques were used to
implement classification task with interesting properties such as provides guidance to the system and
shortening the number of cycles required to learn.
References
[1] Luger and Stubblefield. “Artificial Intelligence” Structures and Strategies for
complex Problem solving, Benjamin / Cummings, Menlo Park, CA, 1998.
[2] Goldberg, David E. “Genetic Algorithm in Search, Optimization, and Machine
Learning”, Addison Wesley Longmont, International Student Edition 1989.
[3] Unemi, Tatsuo,” Scaling up Reinforcement Learning with Human Knowledge as an
Intrinsic Behavior”, Dept. of Information Technology, University of Zurich,
Switzerland unemi@ifi.unizh.ch, Dept. of Information Systems Science, Soka
University, Japan unemi@iss.soka.ac.jp, 1999.
[4] Crook Stamati, ”Evolving expert systems for autonomous agent control using
reinforcement learning”, M.Sc. Thesis, Evolutionary and Adaptive Systems. School of
Cognitive and Computing Sciences. Sussex University. stamati.crook@redware.com,
September 2003.
[5] Pier Luca Lanzi, ”An Introduction to Learning Classifier Systems”, Artificial
Intelligence and Robotics Laboratory, Department di Electronicae Information,
Politecnico di Milano, pierluca.lanzi@polimi.it, 2002.
[6] Sen, Sandip and Weiss, Gerhard, ”Learning in Multiagent Systems”, 1999.
[7] Bacardit Jaume, Ester Bernad´o-Mansilla, and Martin V. Butz, ”Learning Classifier
Systems: Looking Back and Glimpsing Ahead”. ASAP research group, School of
Computer Science, Jubilee Campus, Nottingham, NG8 1BB and Multidisciplinary
Centre for Integrative Biology, School of Biosciences, Sutton Bonington, LE12 5RD,
University of Nottingham, UK, 2009.
[8] Bull Larry, “Learning Classifier Systems: A Brief Introduction”, Faculty of Computing,
Engineering & Mathematical Sciences University of the West of England, Bristol BS16
1QY, U.K. Larry, 2004.
[9] HuntJohn,”learningclassifiersnystems”, Jaydee Technology Ltd, Harthamn Park,
Corsham, Wiltshire, SN13ORP, 2002.
[10] Robert Elliott Smith, Max Kun Jiang, Jaume Bacardit, Michael Stout, Natalio
Krasnogor, Jonathan D. Hirst, ”A learning classifier system with mutual-informationbased fitness”, UK Engineering and Physical Sciences Research Council (EPSRC),2010.
[11] Odetayo Michael O., “On Genetic Algorithms in Machine Learning And Optimisation”,
PhD Thesis, University of Strathclyde, Glasgow, U.K. 1990.
[12] Brownlee Jason, ”Learning Classifier Systems”, Technical Report 070514A,Complex
Intelligent Systems Laboratory, Centre for Information Technology Research, Faculty
of Information and Communication Technologies, Swinburne University of
Technology, Melbourne, Australia, jbrownlee@ict.swin.edu.au, 2007.
[13] Kovacs Tim, ”Advanced topics in machine learnin strength or accuracy Fitness
Evaluation in Learning Classifier Systems”, UNIVERSITY OF BRISTOL, 2002.
[14] Ryan J. Urbanowicz and Jason H. Moore, ”Learning Classifier Systems: A Complet
Introduction, Review, and Roadmap”, Department of Genetics, Dartmouth College,
Hanover, NH 03755, USA, 2009.
[15] Zhou Qing Qing and Purvis Martin, “A Market-Based Rule Learning System” a
Guang Dong Data Communication Bureau China Telecom Dongyuanheng Rd.,
Yuexiunan, 2004.
[16] Nagasaka Ichiro & Kikuchi Makoto & Kitamura. Shinzo, “A Formal Analysis of
Classifier System and Interface Between Learning System and Environment”.
Department of Computer and Systems Engineering, Kobe University, 2002.
[17] Porta Josep M., ”Reinforcement-Based Learning with Automatic Categorization”,
Institut de Robòtica I Informàtica Industrial (CSIC, UPC), Gran Capità 2-4, 08034,
Barcelona (SPAIN), jporta@iri.upc.es, 1999.
[18] Kelly Ian Darrell, ”The Development of Shared Experience Learning in a Group of
Mobile Robots.”, Thesis submitted for the Degree of Doctor of Philosophy Department
of Cybernetics, University of Reading, 1997.
Download all article in PDF
Support the magazine and subscribe to the content
This is premium stuff. Subscribe to read the entire article.