This is an introduction to the theory and practice of computational learning. the author describes in detail both the connectionist techniques ussed in computing with artificial neural networks and the symbolist techniques generally used in artificial intelligence. These techniques are presented in a common framework within which the behaviour of learning mechanisms is analyzed geometrically. This treatment is suitable for the analysis of computational learning, allowing results to be presented in an accessible form using pictures and diagrams. As well as detailed discussion of the principle theories and techniques, problems are given at the end of each chapter and comprehensive reading lists are included. these features make the book particularly suitable for students approaching the subject for the first time.