State of the Art of Reinforcement Learning

Abstract

Yeah so having offered introductions, the focus of this paper will now be to offer a few quick discussions surrounding reinforcement learning domain papers published at next week’s ICLR conference. To keep the project manageable we’ll limit our attention to those papers chosen for spotlight and oral presentations, which is a proxy for selecting papers deemed to have the most significant findings by conference chairs. In some cases the discussions may be more involved than others, this is partly indicative of my limited background in the field as in a few cases may be out of my depth. Hopefully a reader may expect to pick up a few points about current state of the art for the field. Yeah so without further ado.

Tasks

Reproductions