Neural network models. An analysis (Q1906839): Difference between revisions

In the last seven years, I have done research into neural networks, while working and teaching in an engineering environment. I have learned that the knowledge needed to develop successful engineering applications of neural networks, is based on dynamical systems, information theory, and statistical physics. This book uses these three paradigms to explain neural networks. The novelty of this approach is that dynamical systems and information theory concepts are used for the first time in an expository text about neural networks. Where other books try to give an encyclopedic overview, I have chosen a pedagogical approach. Attention is focused on the main issues, and they are developed from scratch. This will allow a wide audience to understand the main issues about neural networks in depth. The reader will learn the most important parameters for the design of neural networks applications. In this way, he or she will start an application with the building blocks ready at hand, without the need for a blind imitation of other existing applications.