This book introduces “functional networks,” a novel neural-based paradigm, and shows that functional network architectures can be efficiently applied to solve many interesting practical problems. Functional networks allow for a more general class of units than the sigmoidal units used in neural networks. They can reproduce some physical or engineering properties; the initial functional network can arise directly from the problem under consideration. Estimation of
functional net parameters can, in many cases, be obtained by solving a linear system of equations. This means a quick and unique solution: the global minimum.
The book is addressed to a wide audience including computer scientists, engineers, mathematicians, etc. No strong prerequisites are assumed, though a previous knowledge of neural networks is convenient. The book consists of 11 chapters grouped in four parts: Neural Networks, Functional Networks, Applications, and Computer Programs. The book also includes an Index. The contents are as follows: “Introduction to Neural Networks,” “Introduction to Functional Networks,” “Functional Equations,” “Some Functional Network Models,” “Model Selection,” “Applications to Time Series,” “Applications to Differential Equations,” “Application to CAD,” “Applications to
Regression,” “Mathematica Programs,” and “A Java Applets”.