RL group discussion Thursday Feb 17, 1:30-2:30, Wean 4601 FOOD What (the !@#$) is the Reinforcement Learning Group at CMU? This discussion is meant to serve several purposes. First, I hope it will enhance the "groupiness" of the group. There are a bunch of new members, and we don't all even know each others' names, much less our research interests. So we can spend a few minutes just eating pizza and shooting the breeze. Second, it will give us a chance to figure out what the scope of the group is. Certainly the name "Reinforcement Learning Group" doesn't fully capture the variety of talks we've had over the last year, ranging from planning, robotics, genetic algorithms and optimization, to function approximation, biology and theory. Sven Koenig has suggested that we consider changing our name to the Graduate AI Seminar, which evidently used to exist here and died out. I prefer a narrower focus, perhaps the "Learning Agents Group", or perhaps sticking with "Reinforcement Learning." But it's for you all to decide. If the discussion goes well, we will end up Thursday with (1) a new group name; (2) knowing each other's names and interests; and (3) a broad but shared set of research questions and approaches which give us group karma. -Justin