Bayesian Automatic Relevance Determination for Feature Selection in Credit Default Modelling

Rendani Mbuvha, Illyes Boulkaibet, Tshilidzi Marwala

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

7 Citations (Scopus)

Abstract

This work develops a neural network based global model interpretation mechanism - the Bayesian Neural Network with Automatic Relevance Determination (BNN-ARD) for feature selection in credit default modelling. We compare the resulting selected important features to those obtained from the Random Forest (RF) and Gradient Tree Boosting (GTB). We show by re-training the models on the identified important features that the predictive quality of the features obtained from the BNN-ARD is similar to that of the GTB and outperforms those of RF in terms of the predictive performance of the retrained models.

Original languageEnglish
Title of host publicationArtificial Neural Networks and Machine Learning – ICANN 2019
Subtitle of host publicationWorkshop and Special Sessions - 28th International Conference on Artificial Neural Networks, Proceedings
EditorsVera Kurková, Igor V. Tetko, Pavel Karpov, Fabian Theis
PublisherSpringer Verlag
Pages420-425
Number of pages6
ISBN (Print)9783030304928
DOIs
Publication statusPublished - 2019
Event28th International Conference on Artificial Neural Networks, ICANN 2019 - Munich, Germany
Duration: 17 Sept 201919 Sept 2019

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11731 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference28th International Conference on Artificial Neural Networks, ICANN 2019
Country/TerritoryGermany
CityMunich
Period17/09/1919/09/19

Keywords

  • Automatic Relevance Determination
  • Bayesian
  • Credit default modelling
  • Hybrid Monte Carlo
  • Neural networks

ASJC Scopus subject areas

  • Theoretical Computer Science
  • General Computer Science

Fingerprint

Dive into the research topics of 'Bayesian Automatic Relevance Determination for Feature Selection in Credit Default Modelling'. Together they form a unique fingerprint.

Cite this