Incremental Learning From Symbolic Objects
This paper deals with discrimination on real-world data when examples as well as partial rules are initially available. As the learning base is both noisy and insufficiently representative, our approach is oriented towards successive approximations of a discriminant rule base; the aim is to predict conclusions according to further examples.
A sub-optimal generalization leads to an approximative rule base including redundancy (strongly overlapping, rules) and errors.
A change of representation called reduction transforms the refinement of the previous rule base into a new learning problem. Iteration of the generalization then defines a second rule set, refining and correcting the previous one, and so on.
KeywordsRule Base Learning Problem Learning Base Incremental Learning Generalization Algorithm
Unable to display preview. Download preview PDF.
- BHATNAGAR, R.K., and KANAL, L.N. (1986), Handling uncertain information: A review of numeric and non-numeric methods, in: Uncertainty in Artificial Intelligence, eds. L.N. Kanal and J.F. Lemmer, North-Holland, Amsterdam.Google Scholar
- BRITO, P., and DIDAY, E. (1990), Pyramidal representation of symbolic objects, this volume.Google Scholar
- CESTNIK, B., BRATKO, I., and KONONENKO, I. (1987), ASSISTANT 86: A knowledge elicitatión tool for sophisticated users, in: Progress in Machine Learning, Proceedings of EWSL 87, eds. I. Bratko and N. Lavrac.Google Scholar
- CLARK, P., and NIBLETT, T. (1987), Induction in noisy domains, in: Progress in Machine Learning, Proceedings of EWSL 87, eds. I. Bratko and N. LavracGoogle Scholar
- DAVIS, L, and STEENSTRUP, M. (1987), Genetic Algorithms and Simulated Annealing: An Overview, in: Genetic Algorithms and Simulated Annealing, ed. L. Davis, Pitman, London.Google Scholar
- DIDAY, E. (1990), Knowledge representation and symbolic data analysis, this volume.Google Scholar
- FALKENHAIMER, B.C., and MICHALSKI, R.S. (1986), Integrating quantitative and qualitative discovery in the ABACUS system, Machine Learning, April.Google Scholar
- HOLLAND, J.H., HOLYOAK, K.J., NISBETT, R.E., and THAGARD, P.R. (1985), Classifier Systems, Q-morphisms, and Induction, in: Genetic Algorithms and Simulated Annealing, ed. L. Davis, Pitman, London.Google Scholar
- KODRATOFF, Y. (1988), Introduction to Machine Learning, Pitman Eds.Google Scholar
- KODRATOFF, Y., and LOISEL, R. (1984), Learning complex structural descriptions from examples, Computer vision, graphics and image processing, 27.Google Scholar
- LANGLEY, P., BRADSHAW, G., and SIMON, H. (1984), Rediscovering chemistry with the BACON system, in: Machine Learning: An Artificial Intelligence Approach, eds. R.S. Michalski, J.G. Carbonnel and T.M. Mitchell, Palo Alto.Google Scholar
- LERMAN, I.-C., LEBBE, J., NICOLAS, J., PETER, P., and VIGNES, R. (1989), Conceptual clustering in biology: applications and perspectives, in: Data Analysis, Learning symbolic and numeric knowledge, ed. E. Diday, Nova Science, Antibes.Google Scholar
- MICHALSKI, R.S. (1984), A theory and methodology for inductive learning, in: Machine Learning: An Artificial Intelligence Approach, eds. R.S. Michalski, J.G. Carbonnel and T.M. Mitchell, Palo Alto.Google Scholar
- MICHALSKI, R.S., DIDAY, E., and STEPP, R.E. (1981), A recent advance in Data Analysis: clustering objects into classes characterized by conjunctive concepts, in: Progress in Pattern Recognition, eds. L.N. Kanal and Rosenfled, North-Holland, Amsterdam.Google Scholar
- MICHALSKI, R.S., MOZETIC, I., HONG, J., and LAVRAC, N. (1986), The AQ15 inductive learning system: an overview and experiments, Proceedings of IMAL, Orsay.Google Scholar
- MITCHELL, T.M. (1982), Generalization as Search, Artificial Intelligence, 18.Google Scholar
- PERRON, M.-C. (1989), Learning differential diagnosis rules from numerical knowledge, in: Data Analysis, Learning symbolic and numeric knowledge, ed. E. Diday, Nova Science, Antibes.Google Scholar
- QUINLAN, J.R. (1986), Induction of decision trees, Machine Learning, 1, 81–106.Google Scholar
- SEBAG, M. (1990), Apprentissage multi-couches: une approche symbolique numérique pour la discrimination à partir d’examples et de règles, Thèse d’Université, Paris-IX Dauphine.Google Scholar
- SEBAG, M., and SCHOENAUER, M. (1988), Generation of rules with certainty and confidence factors from incomplete and incoherent learning bases, Proceedings of EKAW 88, Bonn.Google Scholar
- SEBAG, M., and SCHOENAUER, M. (1989a), Iterative learning and redundant generalizations, in: Data Analysis, Learning numerical and symbolic knowledge, ed. E. Diday, Nova Science, Antibes.Google Scholar
- SEBAG, M., and SCHOENAUER, M. (1989b), Learning rules and meta-rules: towards an incremental learning, Rapport interne, Ecole Polytechnique.Google Scholar
- WILLE, R. (1981), Restructuring lattice theory: an approach based on hierarchies of concepts, Proceedings of the symposium on ordered sets, ed. I. Rival.Google Scholar
- WINSTON, P.H. (1975), Learning structural descriptions from examples, in: The Psychology of Computer Vision, ed. P.H. Winston, McGraw-Hill, New York.Google Scholar