Explaining predictions from a neural network ensemble one at a time

Robert Wall, Pádraig Cunningham, Paul Walsh

Research output: Chapter in Book/Report/Conference proceedingConference contribution

4 Scopus citations

Abstract

This paper introduces a new method for explaining the predictions of ensembles of neural networks on a case by case basis. The approach of explaining individual examples differs from much of the current research which focuses on producing a global model of the phenomenon under investigation. Explaining individual results is accomplished by modelling each of the networks as a rule-set and computing the resulting coverage statistics for each rule given the data used to train the network. This coverage information is then used to choose the rule or rules that best describe the example under investigation. This approach is based on the premise that ensembles perform an implicit problem space decomposition with ensemble members specialising in different regions of the problem space. Thus explaining an ensemble involves explaining the ensemble members that best fit the example.

Original languageEnglish (US)
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Pages449-460
Number of pages12
Volume2431 LNAI
StatePublished - 2002
Externally publishedYes
Event6th European Conference on Principles and Practice of Knowledge Discovery in Databases, PKDD 2002 - Helsinki, Finland
Duration: Aug 19 2002Aug 23 2002

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume2431 LNAI
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other6th European Conference on Principles and Practice of Knowledge Discovery in Databases, PKDD 2002
CountryFinland
CityHelsinki
Period8/19/028/23/02

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Fingerprint Dive into the research topics of 'Explaining predictions from a neural network ensemble one at a time'. Together they form a unique fingerprint.

  • Cite this

    Wall, R., Cunningham, P., & Walsh, P. (2002). Explaining predictions from a neural network ensemble one at a time. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 2431 LNAI, pp. 449-460). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 2431 LNAI).