TY  - CHAP
AB  - Several prominent data-mining studies have evaluated the performance of neural networks (NNs) against traditional statistical methods on the two-group classification problem in discriminant analysis. Although NNs often outperform traditional statistical methods, their performance can be hindered because of failings in the use of training data. This problem is particularly acute when using NNs on smaller data sets. A heuristic is presented that utilizes Mahalanobis distance measures (MDM) to deterministically partition training data so that the resulting NN models are less prone to overfitting. We show this heuristic produces classification results that are more accurate, on average, than traditional NNs and MDM.
VL  - 7
SN  - 978-0-85724-201-3, 978-0-85724-202-0/1477-4070
DO  - 10.1108/S1477-4070(2010)0000007014
UR  - https://doi.org/10.1108/S1477-4070(2010)0000007014
AU  - Smith Gregory E.
AU  - Ragsdale Cliff T.
ED  - Kenneth D. Lawrence
ED  - Ronald K. Klimberg
PY  - 2010
Y1  - 2010/01/01
TI  - A deterministic approach to small data set partitioning for neural networks
T2  - Advances in Business and Management Forecasting
T3  - Advances in Business and Management Forecasting
PB  - Emerald Group Publishing Limited
SP  - 157
EP  - 170
Y2  - 2024/09/24
ER  -