A neural network application for bankruptcy prediction

Abstract

Discusses an application of the back error propagation network for making bankruptcy prediction decisions. Results of simulations with one and two hidden layers with varying nodes are presented. It is observed that the configuration with two hidden layers had a superior classification accuracy compared to the one with a single hidden layer. Based on the initial results it appears that neural network algorithms can be investigated further as potential models for bankruptcy prediction.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>