Marc Métivier

An UCT Approach for Anytime Agent-based Planning

In this paper, we introduce a new heuristic search algorithm based on mean values for anytime planning, called MHSP. It consists in associating the principles of UCT, a bandit-based algorithm which gave very good results in computer games, and especially in Computer Go, with heuristic search in order to obtain an anytime planner that provides partial plans before finding a solution plan, and furthermore finding an optimal plan. The algorithm is evaluated in different classical planning problems and compared to some major planning algorithms. Finally, our results highlight the capacity of MHSP to return partial plans which tend to an optimal plan over the time.

MCTS Experiments on the Voronoi Game

Monte-Carlo Tree Search (MCTS) is a powerful tool in games with a finite branching factor. The paper describes an artificial player playing the Voronoi game, a game with an infinite branching factor. First, it shows how to use MCTS on a discretization of the Voronoi game, and the effects of enhancements such as RAVE and Gaussian processes (GP). Then a set of experimental results shows that MCTS with UCB+RAVE or with UCB+GP are good first solutions for playing the Voronoi game without domain-dependent knowledge. Moreover, the paper shows how the playing level can be greatly improved by using geometrical knowledge about Voronoi diagrams. The balance of diagrams is the key concept. A new set of experimental results shows that a player using MCTS and geometrical knowledge outperforms a player without knowledge.

Planning When Goals Change: A Moving Target Search Approach

Devising intelligent robots or agents that interact with humans is a major challenge for artificial intelligence. In such contexts, agents must constantly adapt their decisions according to human activities and modify their goals. In this paper, we tackle this problem by introducing a novel planning approach, called Moving Goal Planning (MGP), to adapt plans to goal evolutions. This planning algorithm draws inspiration from Moving Target Search (MTS) algorithms. In order to limit the number of search iterations and to improve its efficiency, MGP delays as much as possible triggering new searches when the goal changes over time. To this purpose, MGP uses two strategies: Open Check (OC) that checks if the new goal is still in the current search tree and Plan Follow (PF) that estimates whether executing actions of the current plan brings MGP closer to the new goal. Moreover, MGP uses a parsimonious strategy to update incrementally the search tree at each new search that reduces the number of calls to the heuristic function and speeds up the search. Finally, we show evaluation results that demonstrate the effectiveness of our approach.

Using a Goal-Agenda and Committed Actions in Real-Time Planning

In the context of real-time planning, this paper investigates the contributions of two enhancements for selecting actions. First, the agenda-driven planning enhancement ranks relevant atomic goals and solves them incrementally in a best-first manner. Second, the committed actions enhancement commits a sequence of actions to be executed at the following time steps. To assess these two enhancements, we developed a real-time planning algorithm in which action selection can be driven by a goal-agenda, and committed actions can be done. Experimental results, performed on classical planning problems, show that agenda-planning and committed actions are clear advantages in the real-time context. Used simultaneously, they enable the planner to be several orders of magnitude faster and solution plans to be shorter.