Prediction of financial distress: An application to Chinese listed companies using ensemble classifiers of multiple reductions

Predicting financial distress has been a subject of keen interest in financial economics. In this paper, we forward a financial distress prediction model based on multiple reduction ensembles, which employs neighborhood rough set based attribute reduction to generate a set of reducts, then each redu...

Full description

Saved in:
Bibliographic Details
Published inInternational Conference on Management Science & Engineering ... annual conference proceedings (Print) pp. 1456 - 1461
Main Author Wu, Bao-xiu
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.08.2014
Subjects
Accuracy
Classification algorithms
Companies
ensemble classifiers
financial distress prediction
Logistics
multiple classifier system
neighborhood rough set
Prediction algorithms
Predictive models
Support vector machines
Online AccessGet full text

Cover

More Information
Summary:Predicting financial distress has been a subject of keen interest in financial economics. In this paper, we forward a financial distress prediction model based on multiple reduction ensembles, which employs neighborhood rough set based attribute reduction to generate a set of reducts, then each reduct is used to train a base classifier, and finally their results are combined through simple majority voting. Taking Chinese listed companies' real world data as sample data, adopting 10-fold cross validation technique to assess predictive performance, an experiment study is carried out. By comparing the experiment results with the raw data and the single reduct based classifiers, it is concluded that this model can improve the average prediction accuracy or both accuracy and stability, so it is more suitable for financial distress prediction than the single reduct based classifiers.
ISBN:147995375X
9781479953752
ISSN:2155-1847
DOI:10.1109/ICMSE.2014.6930403