Neural discriminant analysis

In this article the role of the bootstrap is highlighted for nonlinear disc riminant analysis using a feedforward neural network model. Statistical tec hniques are formulated in terms of the principle of the likelihood of a neu ral-network model when the data consist of ungrouped binary responses and a set of predictor variables, We illustrate that the information criterion b ased on the bootstrap method is shown to be favorable when selecting the op timum number of hidden units for a neural-network model, In order to summar ize the measure of goodness-of-fit, the deviance on fitting a neural-networ k model to binary response data can be bootstrapped, We also provide the bo otstrap estimates of the biases of excess error in a prediction role constr ucted by fitting to the training sample in the neural network model. We add itionally propose bootstrap methods for the analysis of residuals in order to identify outliers and examine distributional assumptions in neural-netwo rk model fitting, These methods are illustrated through the analyzes of med ical diagnostic data.