ABSTRACT
Neural network models of machine learning overlap heavily with statistics, espeially nonparametric statistics, since both fields study the analysis of data, but ML is lso an important branch of theoretical computer science. These network models aim t obtaining the best possible generalization performance without making restricting ssumptions in the model on the distributions of data generated by an observed pheomenon. This view of neural networks is adopted in the current chapter. Some of the odels adopted by machine learning originated as nonparametric statistical tools or
xploratory data analysis algorithms. Sometimes the neural network terminology may eem to be redundant to describe some of the methods. Nevertheless, it is traditionally ccepted for historical and methodological reasons.
