Convex Hull Monte-Carlo Tree-Search

Michael Painter, Bruno Lacerda, Nick Hawes

PosterID: 20 PDF Poster BibTeX

This work investigates monte-carlo planning for agents in stochastic (and potentially large) environments, that may have multiple objectives for which the priorities are not known a priori, or may not be easy to quantify. In this work we propose Convex Hull Monte-Carlo Tree-Search, which builds upon Trial Based Heuristic Tree Search and Convex Hull Value Iteration, as a solution to planning with multiple objectives in large environments. Moreover, we consider how to pose the problem of multi-objective planning as a contextual multi-armed bandits problem, giving a principled motivation for how to select actions from the view of contextual regret.

Session E8: Search
Canb 10/28/2020, 00:00 – 01:00
10/29/2020, 17:00 – 18:00
Paris 10/27/2020, 14:00 – 15:00
10/29/2020, 07:00 – 08:00
NYC 10/27/2020, 09:00 – 10:00
10/29/2020, 02:00 – 03:00
LA 10/27/2020, 06:00 – 07:00
10/28/2020, 23:00 – 00:00