Neural nets - their use and abuse for small data sets
Open Access Location
Neural nets can be used for non-linear classification and regression models. They have a big advantage over conventional statistical tools in that it is not necessary to assume any mathematical form for the functional relationship between the variables. However, they also have a few associated problems chief of which are probably the risk of over-parametrization in the absence of P-values, the lack of appropriate diagnostic tools and the difficulties associated with model interpretation. The first of these problems is particularly important in the case of small data sets. These problems are investigated in the context of real market research data involving non-linear regression and discriminant analysis. In all cases we compare the results of the non-linear neural net models with those of conventional linear statistical methods. Our conclusion is that the theory and software for neural networks has some way to go before the above problems will be solved.
Neural nets, Statistical tools, Data sets, Research data, Datasets
Meyer, D. (2000), Neural nets - their use and abuse for small data sets, Research Letters in the Information and Mathematical Sciences, 1, 145-158