A Stochastic View of Optimal Regret through Minimax Duality , COLT, June 2009
A Stochastic View of Optimal Regret through Minimax Duality , COLT, June 2009
Statistics meets Optimization: A new look at multi-armed bandits, Duke Statistics Seminar, April 2009
Stochastic View of Optimal Regret, UPenn Statistics Seminar, April 2009
Beating the Adaptive Bandit with High Probability, Info Theory and Applications, February 2009 [PDF]
Online Learning with Limited Feedback, UPenn CIS, September 2008 [PDF]
Statistics meets Optimization: A new look at the age-old multi-armed bandit problem, UPenn Stats, March 2008
Online Learning with Limited Feedback, MIT, March 2008
Adaptive Online Gradient Descent, NIPS, December 2007 [PDF]
Progress in Online Gradient Descent Methods, IBM Theory of Computation Seminar, November 2007
"Multitask Learning with Expert Advice", COLT, June 2007
"Stability and Consistency of Approximate Empirical Risk Minimization Algorithms", WNAR/IMS Conference, June 2006
"Stability and Consistency of Approximate Empirical Risk Minimization Algorithms", International Workshop on Applied Probability, May 2006
"Stability of Clustering Methods", NIPS 2005 Workshop: Theoretical Foundations of Clustering, December 2005
"Some Applications of Empirical Process Theory in Statistical Learning", UC Berkeley CIS seminar, September 2005
"Some Properties of Empirical Risk Minimization on Donsker Classes", The Fourth International Conference on High Dimensional Probability, June 2005
"Stability Results in Learning Theory", TTI Seminar, April 2005
"Risk Bounds for Mixture Density Estimation", TTI Seminar, March 2005