Controlling the learning process of real-time heuristic search

• Published in: Artificial Intelligence Vol. 146; no. 1; pp. 1 - 41
• Main Authors: Shimbo, Masashi, Ishida, Toru
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.05.2003; Elsevier BV
• Subjects: Adaptive learning; Artificial Intelligence; Convergence process; Rational agent; Real-time heuristic search; Resource-boundedness; Adaptive learning; Convergence process; Rational agent; Resource-boundedness; Real-time heuristic search
• ISSN: 0004-3702; 1872-7921
• DOI: 10.1016/S0004-3702(03)00012-2

Real-time search provides an attractive framework for intelligent autonomous agents, as it allows us to model an agent's ability to improve its performance through experience. However, the behavior of real-time search agents is far from rational during the learning (convergence) process, in that they fail to balance the efforts to achieve a short-term goal (i.e., to safely arrive at a goal state in the present problem solving trial) and a long-term goal (to find better solutions through repeated trials). As a remedy, we introduce two techniques for controlling the amount of exploration, both overall and per trial. The weighted real-time search reduces the overall amount of exploration and accelerates convergence. It sacrifices admissibility but provides a nontrivial bound on the converged solution cost. The real-time search with upper bounds insures solution quality in each trial when the state space is undirected. These techniques result in a convergence process more stable compared with that of the Learning Real-Time A ∗ algorithm.