An Admissible Heuristic Search Algorithm

摘要: This paper introduces an admissible heuristic search algorithm — Search and Learning Algorithm (SLA*). SLA* is developed from the work presented by Korf in Learning-Real-Time-Algorithm (LRTA*). We retain major elements of Korfs LRTA*, improve its performance incorporating a review component to fully reflect effect learning new front states has upon previous states. The combined strategy search, learning, enabled this accumulate knowledge continuously through guided expansion, identify better directions any stage nodes expansion. With assumption non-overestimating initial estimates for all goal, able find optimal solution single problem solving trial with good efficiency. provide proof optimality solution.