Combinatorial prediction games
Jan 25, 2012 4399 views 
en
0.25
0.5
0.75
1.25
1.5
1.75
2
Combinatorial prediction games are problems of online linear optimization in which the action space is a combinatorial space. These games can be studied under different feedback models: full, semi-bandit, and bandit. In first part of the talk we will describe the main known facts about these models and mention some of the open problems. In the second part we will focus on the bandit feedback and describe some recent results which strengthen the link between bandit optimization and convex geometry.