Explanation-based learning: An alternative view

Abstract

Abstract. In the last issue of this journal Mitchell, Keller, and Kedar-Cabelli presented a unifying framework for the explanation-based approach to machine learning. While it works well for a number of systems, the framework does not adequately capture certain aspects of the systems under development by the explanation-based learning group at Illinois. The primary inadequacies arise in the treatment of concept operationality, organization of knowledge into schemata, and learning from observation. This paper outlines six specific problems with the previously proposed framework and presents an alternative generalization method to perform explanation-based learning of new concepts.

Learning logical definitions from relations

“…FOIL finds definitions of relations one at a thne, using other relations as a kind of background knowledge. In this sense, FOIL is much more restricted than explanation-based learners (Mitchell et al, 1986;DeJong and Mooney, 1986) in the way background knowledge can be expressed. Although it would be possible to allow a background relation to be represented by an arbitrary program that determines whether a given constant tuple is in the relation (Pat Langley, Personal Communication, June 1989), this would require considerably more computation than the current algorithm's use of explicit tuples.…”

Case‐Based Planning: A Framework for Planning from Experience

“…The learning done by a case-based planner is learning by remembering. This is a type of learning not addressed either by the theories of concept learning via induction (see Lebowitz, 1980;Michalski & Larson, 1978;Winston, 1970), or explanation-driven learning (see DeJong & Mooney, 1986;Mitchell, Keller, & Kedar-Cabelli, 1986). Case-based planning requires a knowledge-based learning that makes use of the planner's understanding of the world to determine what should be learned and when it should be learned.…”

On-line learning from search failures

“…The learner uses standard explanation-based generalization (Mitchell et al 1986;Dejong & Mooney, 1986) to explain and generalize the failure, taking as its target concept the definition of failure specified by the impossibility theory. First the learner uses this theory to explain why state 7 does not satisfy the current goal (on A B).…”