Composed of three sections, this book presents the training algorithm for neural networks: backpropagation. The first presents the theory and principles, the second a number of network architectures, and the third shows how these can be applied to a number of different fields.