Practical Neural Network Recipies in C++

8

Chapter 1

mathematical discussion of these properties is given in [Hornik, 1991]

and [Blum and Li, 1991]. See page 85 of this book for more details.

The probabilistic neural network also has a solid theoretical

basis.

Under very general conditions, its ability to classify is asymp-

totically Bayes-optimal. See [Specht, 1990] and page 201 of this book

for details.

In summary, many artificial neural networks possess both

substantial theoretical foundations and practical utility. Any problem

that can be solved with traditional modeling or statistical methods can

most likely be solved more effectively with a neural network. It is

hoped that th ...