Stratified case-based reasoning: reusing hierarchical problem solving episodes

摘要: Stratified case-based reasoning is a technique in which abstract solutions produced during hierarchical problem solving are used to assist retrieval, matching, and adaptation. We describe the motivation for integration of with solving, exemplify its benefits, detail set algorithms that implement our approach, present their comparative empirical evaluation on path planning task. Our results show stratified significantly decreases computational expense required retrieve, match, adapt cases, leading performance superior both simple ab initio.