The role of prior causal theories in generalization

摘要: OCCAM is a program which organizes memories of events and learns by creating generalizations describing the reasons for outcomes events. integrates two sources information when forming generalization: • Correlational reveals perceived regularities in events. • Prior causal theories explain events The former has been extensively studied machine learning. Recently, there interest explanation-based learning latter source utilized. In OCCAM, prior are preferred to correlational generalizations. This strategy supported number empirical investigations. Generalization rules used suggest intentional relational relationships. familiar domains, these relationships confirmed or denied differentiate relevant irrelevant features. unfamiliar postulated serve as basis construction theories.