Online Learning
Lorenzo Rosasco/Tomaso Poggio

Description

In this class we are going to introduce online learning from three different perspectives - stochastic approximation, incremental emprical risk minimization and game theoretic. We introduce recursive (incremental) and online algorithms for optimizing RLS, and compare their merits.

Slides

Slides for this lecture: PDF.

Suggested Reading

H. J. Kushner and G. Yin, Stochastic Approximation and Recursive Algorithms and Applications, 2^nd Edition, Springer-Verlag, New York, 2003, [Applications of Mathematics, Volume 35], xxii+474 pp.

D. P. Bertsekas, "Incremental Gradient, Subgradient, and Proximal Methods for Convex Optimization: A Survey", Lab. for Information and Decision Systems Report LIDS-P-2848, MIT, August 2010.

Nicolo Cesa-Bianchi and Gabor Lugosi Prediction, learning, and games Cambridge University Press, 2006.