A General Theoretical Framework for Learning Smallest Interpretable Models

Ordyniak, S. orcid.org/0000-0003-1935-651X, Paesani, G., Rychlicki, M. et al. (1 more author) (2024) A General Theoretical Framework for Learning Smallest Interpretable Models. In: Wooldridge, M., Dy, J. and Natarajan, S., (eds.) Proceedings of the 38th AAAI Conference on Artificial Intelligence. Thirty-Eighth AAAI Conference on Artificial Intelligence, 20-27 Feb 2024, Vancouver, Canada. AAAI Press , Washington, DC, USA , pp. 10662-10669. ISBN 978-1-57735-887-9

Abstract

We develop a general algorithmic framework that allows us to obtain fixed-parameter tractability for computing smallest symbolic models that represent given data. Our framework applies to all ML model types that admit a certain extension property. By showing this extension property for decision trees, decision sets, decision lists, and binary decision diagrams, we obtain that minimizing these fundamental model types is fixed-parameter tractable. Our framework even applies to ensembles, which combine individual models by majority decision.

Metadata

Item Type:	Proceedings Paper
Authors/Creators:	Ordyniak, S. https://orcid.org/0000-0003-1935-651X Paesani, G. Rychlicki, M. Szeider, S.
Editors:	Wooldridge, M. Dy, J. Natarajan, S.
Copyright, Publisher and Additional Information:	© 2024, Association for the Advancement of Artificial Intelligence (www.aaai.org). This is an author produced version of a conference paper published in Proceedings of the AAAI Conference on Artificial Intelligence. Uploaded in accordance with the publisher's self-archiving policy.
Keywords:	KRR: Computational Complexity of Reasoning, ML: Transparent, Interpretable, Explainable ML
Dates:	Published: 24 March 2024 Published (online): 24 March 2024 Accepted: 9 December 2023
Institution:	The University of Leeds
Academic Units:	The University of Leeds > Faculty of Engineering & Physical Sciences (Leeds) > School of Computing (Leeds)
Funding Information:	Funder Grant number EPSRC (Engineering and Physical Sciences Research Council) EP/V00252X/1
Depositing User:	Symplectic Publications
Date Deposited:	19 Jan 2024 14:05
Last Modified:	10 May 2024 14:30
Status:	Published
Publisher:	AAAI Press
Identification Number:	10.1609/aaai.v38i9.28937
Open Archives Initiative ID (OAI ID):	oai:eprints.whiterose.ac.uk:208013

CORE (COnnecting REpositories)

A General Theoretical Framework for Learning Smallest Interpretable Models

Abstract

Metadata

Download

Accepted Version

Export

Statistics