This month we're very lucky to have Rob Schapire who will be giving a talk on learning algorithms for playing repeated games. Rob is a professor at Princeton and is well known in the machine learning community for his work on boosting and all sorts of awesome work in the field. This is an event organzied by the NYC Machine Learning Meetup.

Talk Abstract:

This talk will describe a simple, general algorithm for learning to play any game in a standard matrix form against any unknown adversary.  The algorithm can be shown never to perform much worse than the best fixed strategy, even if selected in hindsight.  Moreover, because of the algorithm's moderate resource requirements, it can be used even when working with extremely large game matrices.  Taken together, these properties make the algorithm a good fit for a range of machine-learning applications, including on-line learning and boosting.  Recently, the algorithm has also been applied to reinforcement learning, specifically, to the problem of learning to imitate the behavior of an "expert" while attempting simultaneously to improve on the expert's performance.