Topic: Description:Model based search methods and their application to RL
The objective of the talk is to introduce a particular class of zero-order optimization methods called model-based search methods which are gradient-free techniques to generate high quality solutions to the optimization (deterministic or stochastic) problem. These methods have the unique characteristic to operate in a black-box setting where it is presumed that the analytic closed form expression of the objective function is unavailable and hence they are completely non-dependent on the structural properties of the objective function. In this talk, I explore various algorithms of this class, their efficient extensions and their application to reinforcement learning.
The Tea Time Talks are a series of talks primarily given by the students and faculty studying Artificial Intelligence at the University of Alberta, and provide a comfortable, informal space in which to listen and learn about topics pertaining to machine intelligence and machine learning.