I got the opportunity to present my recent work on turn-taking prediction on 2016 AAAI symposium on AI-HRI. It was such a great event which features some interesting events, like breakout session (to discuss hot topics with your peers) and debut-like keynotes. I got the opportunity to make lots of academic friends.
Some comments and questions that I got about my talk (full talk recording here):
This blog (and github) shows a very easy example of using Deep Reinforcement Learning (DRL) to play the game Catch. It provides the code to start the game simulation environment (very simple) and the associated code for training and testing DRL. It is an excellent toy example to start playing with DRL. In the following section, I will record my notes and comments about playing with this piece of code.
Train and Test
You need to know who are the best experts in DRL right now, This is by no means a full list and I just got started viewing this area.