- Title: Evolutionary MCTS for Game AI and Beyond
- Speaker: Hendrik Baier, CWI Amsterdam
- Time and date: 11 am to 12 pm, Apr 30, 2019
- Room: BR3.02, Bancroft Road Teaching Rooms
On Tuesday 30th April 2019 the Game AI Group will host a seminar by Hendrik Baier from CWI Amsterdam. All welcome (especially students), no pre-booking required.
Abstract
In this talk, I’m going to present a research direction that tries to combine the strengths of Monte Carlo Tree Search with those of Evolutionary Algorithms. Summarizing work presented at CIG and AIIDE last year, I’m first going to explain how Evolutionary MCTS (EMCTS) works, apply it to online planning in turn-based tactics/strategy games, and show how it can be extended to include a simple opponent model. Afterwards, I’m going to briefly go beyond the goal of “strongest possible AI” by showing first results on dynamic difficulty adaptation with EMCTS; and beyond the field of game AI altogether by discussing how I am applying RHEA, MCTS, and EMCTS in the context of managing flexibility on future energy markets!
Bio
I am currently living in the beautiful city of Amsterdam, working as researcher in the Intelligent and Autonomous Systems group at CWI. Before, I worked as research associate for artificial intelligence and data analytics at Digital Creativity Labs in York, England, and as research fellow in artificial intelligence in the Advanced Concepts Team of the European Space Agency in Noordwijk, Netherlands. In 2015, I finished my Ph.D. in the Games and AI group at the Department of Knowledge Engineering, Maastricht University. My topic was planning and search — the search for an optimal strategy in a given problem, or just for a good next action to take. I used games as model problems. Application areas are both adversarial and collaborative scenarios with several agents, as well as optimization problems with a single agent.