Abstract

Any learning process can be viewed as a self-modification of the leaxnefs current knowledge tArough an. interaction with some information source. Such knowledge modification is guided by the learner's deshe to achieve a certain outcome, and can engage any kind of inference. The type of inference involved depends on he input information, the current (background) knowledge and the learneFs task ax hand. Based on such a view of learning, several fundamental concepts are analized and clarified, in paxticular, analytic and synthetic learning, derivm:ional and hypothetical explanation, constnictive induction, abduction, abstraction and deductive generalization. It is shown that inductive generalization and abduction can be viewed as two basic forms of general induction, and that abstraction and deductive generalization axe two related forms of constructive deduction. Using this conceptual framework, a methodology for multistrategy task-adaptive learning (MTL) is outlined, in which learning strategies axe combined dynamically, depending on the current learning situation. Speccally, an MTL learner anaLizes a "wiad" relationship among the input information, the background knowledge and the learning task, and on that basis determines which strategy, or. a combination thereof, is most appropriate at a given learning step. To implement the MTL methodology, a new knowledge representation is proposed, based on the parametric association rules (PARs). Basic ideas of MTL are illustrated by means of the well-known "cup" example, through which is shown how an MTL learner can employ, depending the above mad relationship, emprical learning, constructive inductive generalization, abduction, explanation-based learning and absuaction.

Citations

6071	Genetic Algorithms - Goldberg - 1989
333	Explanation-Based Learning: An Alternative View - DeJong, Mooney - 1986
284	Qualitative Results Concerning the Utility of Explanation-Based Learning - Minton - 1990
272	An assumption-based truth maintenance system - Kleer - 1986
246	Chunking in Soar: The anatomy of a general learning mechanism - Laird, Rosenbloom, et al. - 1986
180	A General Lower Bound on the Number of Examples Needed - Ehrenfeucht, Haussler, et al. - 1989
68	Learning at the knowledge level - Dietterich - 1986
64	Causes for events: their computation and applications - Cox, Pietrzykowski - 1986
61	Incremental learning from noisy data - Schlimmer, Granger - 1986
50	Protos: An exemplar-based learning apprentice - Bareiss, Porter - 1987
38	Understanding the nature of learning: issues and research directions - Michalski - 1986
31	Learning flexible concepts: Fundamental ideas and a method based on two-tiered representaion - Michalski - 1990
31	Discovering rules from large collections of examples: A case study,” Expert Systems - Quinlan - 1979
27	Acquiring Effective Search Control Rules: Explanation-Based Learning - Minton, Carbonell, et al. - 1987
25	Genetic Algorithms and Classifier Systems: Foundations and Future Directions - Holland - 1987
25	Integrated learning: Controlling explanation - Lebowitz - 1986
23	Apprenticeship learning in imperfect domain theories - Tecuci, Kodratoff - 1990
21	Discovering Classification Rules Using Variable-Valued Logic System VL 1 - Michalski - 1973
18	The Use of Explanations for Similarity-Based Learning - Danyluk - 1987
18	Integrating Explanation-Based and Empirical Learning Methods in OCCAM - Pazzani - 1988
14	INDUCE.4: A Program for Incrementally Learning Structural Descriptions from Examples - Bentrup, Mehler, et al. - 1987
12	An Overview of the Odysseus Learning Apprentice - Wilkins, Clancey, et al. - 1986
11	INDUCE 2: A Program For Learning Structural Descriptions From Examples - Hoff, Michalski, et al. - 1983
11	DISCIPLE-1: Interactive apprentice system in weak theory elds - Y, Tecuci - 1987
9	Structural learning in connectionist systems - Barto, Andersen - 1985
8	Connectionist AI, symbolic AI, and the brain - Smolensky - 1987
7	An experiment on inductive learning in chess end games - Michalski, Negri - 1977
6	Analytics, in The Works of Aristotle, Volume 1 - Aristotle - 1987
6	Recent Results in the Use of Context for Learning New Rules - P - 1989
6	Processing Issues in Comparisons of Symbolic and Connectionism Learning Systems - Fisher, McKusick, et al. - 1989
6	Theory formation by abduction: Initial results of a case study based on the chemical revolution - O’Rorke, Morris, et al. - 1989
6	Explanation-Based Learning of Generalized Robot Assembly Plans - Segre - 1987
4	Learning to predict sequences.” In Machine learning: an artificial intelligence approach II, edited by - Dietterich, Michalski - 1986
4	A Tool for the Management of Incomplete Theories: Reasoning about Explanations - Duval, Kodraxoff - 1990
4	Constructive Closed-loop Learning: Fundamental Ideas and Examples, a paper submitted for presentation at - Michalski, Watanabe
4	Elements of Logic, in Collected Papers of Charles Sanders Peirce (1839 1914), Ch - Peirce - 1965
3	F_xts.) The Great Ideas: A Synoptic of Great - Adler, Gorman - 1987
3	Assembly Planning Through Constructive Learning Approac.h," to appear - Ko - 1990
3	Types of Explanation and Their Role n Constructive Learning." to appear in Machine Learning and Inference Reports - Ko, Michalski - 1990
2	Analytics, in Great Books of the - Aristotle - 1987
2	An Approach to Learning from Observation," Chapter 19 in this book - DeJong - 1985
2	An Inductive Approach to Solving the Imperfect Theory Problem,"Proceedings of 1988 Symposium on Explanation-Based Learning - Dieuerich, laran, et al. - 1988
2	Qualitative Prediction: The SPARC/G Methodology for Describing and Predicting Discrete Processes - Michalski, Ko, et al. - 1987
2	Explaining Kowledge Systems:Justifying Diagnostic Conclusions - Tanner - 1990
1	Program Synthesis as a Theory Formation Tk: Problem Representations and Solution Methods - Amarel - 1986
1	An Overview of - Mitchell - 1983
1	The Logic of Plausible Reasoning: A Cor Theory - Collins, Michalski - 1989
1	Genetic Algorithrn and Simulated Annealing - Davis - 1987
1	Toward a unified theory of learning: Multistrategy task-ada12five learning - DeJong - 1988
1	Using Genetic Algorithms to Solve NP-Complete Problems - long, Spea - 1989