Approaches for Rule Discovery in a Learning Classifier System

Michael Heider, Helena Stegherr, David Pätzel, Roman Sraj, Jonathan Wurth, Benedikt Volger, Jörg Hähner

2022

Abstract

To fill the increasing demand for explanations of decisions made by automated prediction systems, machine learning (ML) techniques that produce inherently transparent models are directly suited. Learning Classifier Systems (LCSs), a family of rule-based learners, produce transparent models by design. However, the usefulness of such models, both for predictions and analyses, heavily depends on the placement and selection of rules (combined constituting the ML task of model selection). In this paper, we investigate a variety of techniques to efficiently place good rules within the search space based on their local prediction errors as well as their generality. This investigation is done within a specific LCS, named SupRB, where the placement of rules and the selection of good subsets of rules are strictly separated in contrast to other LCSs where these tasks sometimes blend. We compare a Random Search, (1,λ)-ES and three Novelty Search variants. We find that there is a definitive need to guide the search based on some sensible criteria, i.e. error and generality, rather than just placing rules randomly and selecting better performing ones but also find that Novelty Search variants do not beat the easier to understand (1,λ)-ES.

Paper Citation

in Harvard Style

Heider M., Stegherr H., Pätzel D., Sraj R., Wurth J., Volger B. and Hähner J. (2022). Approaches for Rule Discovery in a Learning Classifier System. In Proceedings of the 14th International Joint Conference on Computational Intelligence (IJCCI 2022) - Volume 1: ECTA; ISBN 978-989-758-611-8, SciTePress, pages 39-49. DOI: 10.5220/0011542000003332

in Bibtex Style

@conference{ecta22,
author={Michael Heider and Helena Stegherr and David Pätzel and Roman Sraj and Jonathan Wurth and Benedikt Volger and Jörg Hähner},
title={Approaches for Rule Discovery in a Learning Classifier System},
booktitle={Proceedings of the 14th International Joint Conference on Computational Intelligence (IJCCI 2022) - Volume 1: ECTA},
year={2022},
pages={39-49},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0011542000003332},
isbn={978-989-758-611-8},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 14th International Joint Conference on Computational Intelligence (IJCCI 2022) - Volume 1: ECTA
TI - Approaches for Rule Discovery in a Learning Classifier System
SN - 978-989-758-611-8
AU - Heider M.
AU - Stegherr H.
AU - Pätzel D.
AU - Sraj R.
AU - Wurth J.
AU - Volger B.
AU - Hähner J.
PY - 2022
SP - 39
EP - 49
DO - 10.5220/0011542000003332
PB - SciTePress