Application of feature extractive algorithm to bankruptcy prediction

Author

Charalambous, Chris ; Charitou, Andreas ; Kaourou, Froso

Author_Institution

Dept. of Bus. Admin., Cyprus Univ., Nicosia, Cyprus

Volume

5

fYear

2000

fDate

2000

Firstpage

303

Abstract

This study uses the feature selection algorithm proposed by Setiono and Liu (1997) to select the most relevant features for the bankruptcy prediction problem. The method uses a feedforward neural network with one hidden layer to decide which features to be removed. Our data consists of 139 matched pair of bankrupt and nonbankrupt US firms for the period 1983-1994. The results of this study indicate that the final neural network obtained with reduced number of inputs gives significantly better prediction results than the one that uses all initial features

Keywords

business data processing; feedforward neural nets; financial data processing; multilayer perceptrons; bankruptcy prediction; feature extraction algorithm; feature selection algorithm; feedforward neural network; hidden layer; two layer neural net; Accuracy; Costs; Economic forecasting; Feature extraction; Feedforward neural networks; Neural networks; Prediction algorithms; Principal component analysis; Testing; Training data;

fLanguage

English

Publisher

ieee

Conference_Titel

Neural Networks, 2000. IJCNN 2000, Proceedings of the IEEE-INNS-ENNS International Joint Conference on

Conference_Location

Como

ISSN

1098-7576

Print_ISBN

0-7695-0619-4

Type

conf

DOI

10.1109/IJCNN.2000.861479

Filename

861479