Pilot Study on the Localized Generalization Error Model for Single Layer Perceptron Neural Network

We had developed the localized generalization error model for supervised learning with minimization of mean square error. In this work, we extend the error model to single layer perceptron neural network (SLPNN). For a trained SLPNN and a given training dataset, the proposed error model bounds above the error for unseen samples which are similar to the training samples. This pilot study is the important first step of investigating localized generalization error models for multilayer perceptron neural networks and support vector machines with sigmoid kernel functions. The characteristics of the error model for SLPNN and how to compare SLPNNs´ generalization capabilities using the error model are also discussed in this paper