Statistical Learning Theory,
Vladimir Vapnik, John Wiley & Sons, 1998

A comprehensive look at the statistical theory of learning and generalization from one of the giants in the field. The book covers the whole picture of learning and generalization theory, presenting a method for determining necessary and sufficient conditions for consistency in the learning process and coverin function estimates from small data pools and their application toreal-lif problems. These theories have applications in many fields, including psychology, education, computer science and robotics.