Elad Hazan: Is Optimization Computationally Equivalent to Online Learning?

Tuesday, May 12, 2015 - 4:15pm to 5:15pm
Refreshments: 
Light Refreshments at 4pm
Location: 
G449 (Patil/Kiva)
Speaker: 
Elad Hazan

Abstract: 

Vapnik's fundamental theorem of statistical learning establishes a
computational equivalence between optimization (Empirical Risk
Minimization) and learning in the statistical setting. Is the same
true for learning in games? We give a precise answer to this question.

Joint work with Tomer Koren