作者: Prasad Tadepalli
DOI:
关键词:
摘要: Speedup learning seeks to improve the efficiency of search-based problem solvers. In this paper, we propose a new theoretical model speedup which captures systems that solving performance by user-given set problems. We also use motivate notion "batch solving," and argue it is more congenial than sequential solving. Our results are applicable all serially decomposable domains. empirically validate our in domain Eight Puzzle.