Topics in Statistical Learning Theory (Panoramas Et Syntheses, 57)

This volume is the outcome of a series of three lectures on statistical learning theory given at Institute Henri Poincaré in 2011 under the auspices of the Société Mathéatique de France. The introductory chapter provides an overview of the history of Statistical Learning Theory, its roots, and its mathematical tools. The chapter Algorithms for minimally supervised learning, by Sanjoy Dasgupta, describes the progress of theoretical computer science on the issues of unsupervised learning (clustering) and active learning. Surprisingly, much of this progress is due to the confrontation of measurement concentration theory, complexity theory, and established practices in numerical statistics. The chapter Online prediction, by Peter Bartlett, focuses on online learning. It is a confrontation between statistics, game theory and optimization. Information for our distributors: A publication of the Société Mathématique de France, Marseilles (SMF), distributed by the AMS in the U.S., Canada, and Mexico. Orders from other countries should be sent to the SMF. AMS individual members receive a 10% discount and members of the SMF receive a 30% discount from list. No other discounts apply.