IRMA-International.org: Creator of Knowledge
Information Resources Management Association
Advancing the Concepts & Practices of Information Resources Management in Modern Organizations

Training Distribution Strategies for Optimizing Neural Network Classification Models

Training Distribution Strategies for Optimizing Neural Network Classification Models
View Free PDF
Author(s): Steven Walczak (University of Colorado at Denver, USA), Irena Yegorova (City University of New York, USA)and Bruce H. Andrews (University of Southern Maine, USA)
Copyright: 2002
Pages: 3
Source title: Issues & Trends of Information Technology Management in Contemporary Organizations
Source Editor(s): Mehdi Khosrow-Pour, D.B.A. (Information Resources Management Association, USA)
DOI: 10.4018/978-1-930708-39-6.ch209
ISBN13: 9781930708396
EISBN13: 9781466641358

Abstract

Neural networks have been repeatedly shown to outperform traditional statistical modeling techniques for both discriminant analysis and forecasting. While questions regarding the effects of architecture, input variable selection, learning algorithm, and size of training sets on the neural network model’s performance have been addressed, little attention has been focused on distribution effects of training and out-of-sample populations on neural network performance. This article examines the effect of changing the population distribution within training sets, in particular for a credit risk assessment problem.

Body Bottom