Evaluating the effectiveness of Bayesian feature selection

Research output: Chapter in Book/Report/Conference proceedingChapter

View graph of relations Save citation

Authors

Research units

Abstract

A practical Bayesian approach for inference in neural network models has been available for ten years, and yet it is not used frequently in medical applications. In this chapter we show how both regularisation and feature selection can bring significant benefits in diagnostic tasks through two case studies: heart arrhythmia classification based on ECG data and the prognosis of lupus. In the first of these, the number of variables was reduced by two thirds without significantly affecting performance, while in the second, only the Bayesian models had an acceptable accuracy. In both tasks, neural networks outperformed other pattern recognition approaches.

Request a copy

Request a copy

Details

Publication date2005
Publication titleApplications of probabilistic modelling in medical informatics and bioinformatics
EditorsR. Dybowski, D. Husmeier, S. J. Roberts
PublisherSpringer
Pages371-389
Number of pages19
ISBN (Print)9781846281198
Original languageEnglish

Publication series

NameAdvanced information and knowledge processing
PublisherSpringer Verlag

    Keywords

  • Bayesian approach, inference, neural network models, diagnostic tasks, heart arrhythmia, prognosis of lumps

DOI

Employable Graduates; Exploitable Research

Copy the text from this field...