Managerial Applications of Neural Networks: The Case of Bank Failure Predictions

Abstract

This paper introduces a neural-net approach to perform discriminant analysis in business research. A neural net represents a nonlinear discriminant function as a pattern of connections between its processing units. Using bank default data, the neural-net approach is compared with linear classifier, logistic regression, kNN, and ID3. Empirical results show that neural nets is a promising method of evaluating bank conditions in terms of predictive accuracy, adaptability, and robustness. Limitations of using neural nets as a general modeling tool are also discussed.neural networks, artificial intelligence, discriminant analysis, bank failure predictions

VIP-targeted CRM strategies in an open market

“…Though it is really difficult to make a universally approved definition of CRM, it can be explained as a comprehensive strategy for acquiring, retaining and partnering with selective customers to create value for both a company and its customers. Many previous CRM-related researches have applied data mining techniques to analyse and understand customer behaviours and characteristics, and most of them have worked well (Bortiz et al ., 1995;Fletcher et al ., 1993;Langley et al ., 1995;Lau et al ., 2003;Salchenberger et al ., 1992;Su et al ., 2002;Tam et al ., 1992;Zhang et al ., 1999). In this section, we review previous researches mainly on classification and association rules for a variety of tasks in CRM domain.…”

Predicting bankruptcy using neural networks and other classification methods: The influence of variable selection techniques on model accuracy

“…Tam and Kiang [72] Variables that belong to the following financial categories: asset, income, liquidity…”

“…? 74.10% Tam and Kiang [72] 85.00% 89.00% 87.00% Tan and Dihardjo [73] 92.35% 64.70% 92.23% Tung et al [74] 99.31% 54.54% 73.12% Tyree and Long [33] 100.00% 94.55% 97.95% Vieira et al [75] ? ?…”

A training algorithm for classification of high-dimensional data

“…Although this is a di cult topic, it is of extreme importance for the company's shareholders. Traditionally carried out by accounting experts using heuristic rules, lately this problem has also been tackled by automatic methods, based on statistical and empirical analyses, or adaptive techniques such as neural networks [15]. In this case the sample consists of 450 non-ÿnancial companies, half of which boasted of good ÿnancial health and the other half had failed.…”