Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Toward optimal classifier system performance in non-Markov environments.

P L Lanzi1, S W Wilson

  • 1Dip. di Elettronica e Informazione, Politecnico di Milano, Italy. lanzi@elet.polimi.it

Evolutionary Computation
|December 29, 2000
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

The use of venous anastomotic flow couplers to monitor buried free DIEP flap reconstructions following nipple-sparing mastectomy.

JPRAS open·2020
Same author

Distribution of Pax6 protein during eye development suggests discrete roles in proliferative and differentiated visual cells.

Development genes and evolution·2016
Same author

Management of carpal tunnel syndrome in renal dialysis patients using an extended carpal tunnel release procedure.

Journal of plastic, reconstructive & aesthetic surgery : JPRAS·2007
Same author

Use of mitek anchors in the digits.

Techniques in hand & upper extremity surgery·2006
Same author

Hedgehog and Fgf signaling pathways regulate the development of tphR-expressing serotonergic raphe neurons in zebrafish embryos.

Journal of neurobiology·2004
Same author

The Mexican hat splint--a new splint for the treatment of closed mallet finger.

Journal of hand surgery (Edinburgh, Scotland)·2001
Same journal

Computing Optimal Populations for Binary Problems using Logic Minimization.

Evolutionary computation·2026
Same journal

Enhancing Generalization and Scalability for Multi-Objective Optimization with Population Pre-Training.

Evolutionary computation·2026
Same journal

XCS for Sequential Perceptual Aliasing in Multi-Step Decision Making.

Evolutionary computation·2026
Same journal

A dynamic multi-objective evolutionary algorithm using dual-space prediction and surrogate-based sampling.

Evolutionary computation·2026
Same journal

Adapting MOEA/D to CMA-ES for Dealing with Ill-conditioned Multiobjective Problems.

Evolutionary computation·2026
Same journal

Editorial of the Special Issue: Parallel Problem Solving from Nature PPSN 2024 Extended Versions of Best Paper Candidates.

Evolutionary computation·2026
See all related articles

The XCS classifier system, using Wilson's bit-register memory, achieved near-optimal performance in non-Markov environments with two key extensions: deterministic internal action selection and an oversized register for aliasing. This research enhances reinforcement learning in complex environments.

Area of Science:

  • * Computational intelligence
  • * Machine learning
  • * Reinforcement learning

Background:

  • * Wilson's (1994) bit-register memory scheme is a foundational concept in classifier systems.
  • * Classifier systems are often evaluated in Markov environments, but performance in non-Markov settings presents unique challenges.
  • * Environmental aliasing, where different states are indistinguishable, can impede learning.

Purpose of the Study:

  • * To investigate the efficacy of Wilson's bit-register memory scheme within the XCS classifier system.
  • * To explore modifications for improving performance in non-Markov environments.
  • * To analyze the impact of specific extensions on learning and exploration.

Main Methods:

  • * Implementation of Wilson's bit-register memory scheme into the XCS classifier system.

Related Experiment Videos

  • * Testing the system in a series of non-Markov environments.
  • * Introducing and evaluating two extensions to the original memory scheme: probabilistic external action exploration with deterministic internal action (register setting) selection, and the use of an oversized register.

    • Main Results:

    • * The XCS classifier system demonstrated near-optimal performance in challenging non-Markov environments.
    • * The first extension, deterministic internal action selection, was crucial for effective exploration.
    • * The second extension, employing a register with excess bit-positions, effectively addressed environmental aliasing.

    Conclusions:

    • * The investigated extensions significantly enhance the XCS classifier system's ability to perform in non-Markov environments.
    • * Deterministic internal action selection and oversized registers are effective strategies for overcoming limitations in complex environments.
    • * This work provides valuable insights into adapting reinforcement learning agents for non-Markovian settings.