Quantitative results concerning the utility of explanation-based learning

Minton, Steven

doi:10.1016/0004-3702(90)90059-9

Cited by 281 publications

(213 citation statements)

References 10 publications

Supporting

Mentioning

211

Contrasting

Order By: Relevance

“…The "utility problem" occurs when additional knowledge learned decreases a reasoning system's performance instead of increasing it [10,13]. Theoretically this will always occur in a CBR system when the system's case base increases without bound.…”

Section: Related Researchmentioning

confidence: 99%

An Efficient Hybrid Classification Algorithm – An Example from Palliative Care

Houeland

Aamodt

2011

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. In this paper we present an efficient hybrid classification algorithm based on combining case-based reasoning and random decision trees, which is based on a general approach for combining lazy and eager learning methods. We use this hybrid classification algorithm to predict the pain classification for palliative care patients, and compare the resulting classification accuracy to other similar algorithms. The hybrid algorithm consistently produces a lower average error than the base algorithms it combines, but at a higher computational cost.

show abstract

Section: Related Researchmentioning

confidence: 99%

An Efficient Hybrid Classification Algorithm – An Example from Palliative Care

Houeland

Aamodt

2011

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…Indiscriminate macro learning can degrade problem-solving performance due to the cost of testing the applicability conditions of too many learned macros (Minton, 1985). Past research has investigated heuristics for restricting when to learn macros (Iba, 1989) and which learned macros to retain (Minton, 1990). It should be emphasized that FAILSAFE-2 learns macros only as exceptions to overgeneral censors, tests their applicability conditions only when the original overgeneral censor would have applied, and does not apply all macros in all states, just the ones whose perconditions are fortuitously found to hold while testing the censors, and which are therefore known to achieve the current goal.…”

Section: Learning Macros From Exceptionsmentioning

confidence: 99%

“…All this less directly related work lies outside the scope of this article; for pointers to some of it, see Bhatnagar (1992). FAILSAFE-2 is most closely related to PRODIGY (Minton, 1990) (and its descendants ULS (Chase et at., 1989) and STATIC (Etzioni, 1991)) in its explanation-based learning approach to the acquisition of search control knowledge for a problem solver. We now discuss three key differences from them--the type of knowledge acquired, the technique used to learn it.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

On-line learning from search failures

Bhatnagar¹,

Mostow

1994

Mach Learn

View full text Add to dashboard Cite

Abstract. Learning by explaining failures and avoiding similar ones thereafter is an attractive way to speed up problem solving. However, previous methods for explanation-based learning from failure can take too long to detect failures, explain them, or test the learned rules. This expense is especially critical for adaptive search, in which control knowledge acquired while solving an individual problem instance must be learned quickly enough to speed up its solution.We present an adaptive search technique that speeds up state-space search by learning heuristic censors while searching. The censors speed up search by pruning away more and more of the space until a solution is found in the pruned space. Censors are learned by explaining dead ends and other search failures. To learn quickly, the technique overgeneralizes by assuming that certain constraints are preservable, i.e., remain true along at least one solution path. A recovery mechanism detects violations of this assumption and selectively relaxes learned censors. The technique, implemented in an adaptive problem solver named FAILSAFE-2, learns useful heuristics that cannot be learned by other reported methods.We present experimental evidence that FAILSAFE-2 is effective (learns useful rules, even in recursive domains where PRODIGY and STATIC do not), adaptive (learns fast enough to pay off even within a single problem), and general (speeds up diverse problem solvers, even initially strong ones).

show abstract

“…Minton's work (1985;1988b;1988a) is a particularly notable contribution to this topic. At the same time, there has been a growing awareness of the many ways in which even correct knowledge can be harmful (see for example Markovitch and Scott (1988), Tambe and Newell (1988), and Wilkins and Ma (1989).…”

Section: Introductionmentioning

confidence: 99%

Information filtering: Selection mechanisms in learning systems

Markovitch

Scott

1993

Mach Learn

View full text Add to dashboard Cite

Editor: Bruce PorterAbstract. Knowledge has traditionally been considered to have a beneficial effect on the performance of problem solvers but recent studies indicate that knowledge acquisition is not necessarily a monotonically beneficial process, because additional knowledge sometimes leads to a deterioration in system performance. This paper is concerned with the pmbtem of harmful knowledge: that is, knowledge whose removal would improve a system's performance.In the first part of the paper a unifying framework, called the information filtering model, is developed to define the various alternative methods for eliminating such knowledge from a learning system where selection processes, called filters, may be inserted to remove potentially harmful knowledge. These filters are termed selective experience, selective attention, selective acquisition, selective retention, and selective utilization. The framework can be used by developers of learning systems as a guide for selecting an appropriate filter to reduce or eliminate harmful knowledge.In the second part of the paper, the framework is used to identify a suitable filter for solving a problem caused by the acquisition of harmful knowledge in a learning system called LAssY. LASSY is a system that improves the performance of a PROLOG interpreter by utilizing acquired domain specific knowledge in the form of lemmas stating previously proved results. It is shown that the particular kind of problems that arise with this system are best solved using a novel utilization filter that blocks the use of lenmaas in attempts to prove subgoals that have a high probability of failing.

show abstract

Quantitative results concerning the utility of explanation-based learning

Cited by 281 publications

References 10 publications

An Efficient Hybrid Classification Algorithm – An Example from Palliative Care

An Efficient Hybrid Classification Algorithm – An Example from Palliative Care

On-line learning from search failures

Information filtering: Selection mechanisms in learning systems

Contact Info

Product

Resources

About