Abstract
In this paper the current approach to automatic concept acquisition is criticized. In particular, it is argued that using only generalization, coupled with a simplicity criterion for selecting hypotheses, generates concept descriptions which are poor with respect to their information content, difficult to agree upon by humans and strictly task dependent.
A new framework is proposed instead, based on notion of abstraction, in the sense of Plaisted and Tenenberg. The novelty of this paper, with respect to previous ones, occasionally mentioning abstraction in machine learning, is that it gives a precise definition of abstraction, shows its relation with generalization and also offers a computational framework. Moreover, a special type of abstraction, particularly useful for the learning task, is defined, and an algorithm for computing it is also presented. This type of abstraction can be applied to a body of background knowledge (domain theory), allowing EBG to be performed in a more synthetic representation space. This transformation can (at least partially) offer a solution to the problem of domain theory intractability.
A complete example of domain theory abstraction is also worked out, for the sake of exemplification.
Chapter PDF
References
Amsterdam J. Some Philosophical Problems with Formal Learning Theories. In Proc. AAAI-88, pp. 580–584, 1989
Bergadano F., Giordana, G. & Saitta L. (1988). Automated Concept Acquisition in Noisy Environments. IEEE Trans. on Pattern Analysis and Machine Intelligence, 10, 555–578, 1988.
Buntine W. Generalized Subsumption and its Applications to Induction and Redundancy. Artificial Intelligence, 36, 149–176, 1988.
Carbonell J. (Ed). Artificial Intelligence. Special Issue on Machine Learning, 40, 1989.
Clark K.L. Negation as failure. In Gallaire H. and Minker J. (Eds) Logic and Databases, Plenum Press, New York, pp. 293–372, 1978.
Dejong G. & Mooney R. Explanation-Based Learning: An Alternative View. Machine Learning, 1, 145–176, 1986.
Drastal G., Czako G. & Raatz S. Induction in an Abstraction Space. Proc. IJCAI-89, Detroit, MI, pp. 708–712, 1989.
Giordana A., Roverso D. & Saitta L. Abstraction: An Alternative Approach to Concept Acquisition. In Proceedings of ISMIS-90, pp. 379–387, Knoxville, TN, 1990.
Giordana A. & Saitta L. Abstraction: A General Framework for Learning. In Proceedings of AAAI Workshop on Automatic Generation of Abstraction and Approximation, pp. 245–256, Boston, MA, 1990
Iba W., Wogulis J. & Langley P. Trading-off Simplicity and Coverage in Incremental Concept Learning. Proc. 5th Int. Conf. on Machine Learning, Ann Arbor, MI, pp. 73–79, 1988.
Knoblock C. A Theory of Abstraction for Hierachical Planning. in D.P. Benjamin (Ed.), Change of Representation and Inductive Bias, Kluwer Publ. Co., Boston, MA, 1989a.
Knoblock C. Learning Hierarchies of Abstraction Spaces. Proc. 6th Int. Workshop on Machine Learning Ithaca, NY, 1989b.
Matheus C. & Rendell L. Constructive Induction in Decision Trees. In Proc. IJCAI-89, Detroit, MI, pp. 645–650, 1989.
Michalski R. A Theory and Methodology of Inductive Learning. In Michalski R., Mitchell T. & Carbonell J. (Eds), pp. 83–134, Morgan Kaufmann, Palo Alto, CA, 1983.
Michalski R., Carbonell J. & Mitchell T. (Eds). Machine Learning: An AI Approach, Vol. I. Morgan Kaufmann, Los Altos, CA, 1983.
Michalski R., Carbonell J. & Mitchell T. (Eds). Machine Learning: An AI Approach, Vol. II. Morgan Kaufmann, Los Altos, CA, 1985.
Michalski R. & Kodratoff Y. (Eds). Machine Learning: an Artificial Intelligence Approach, Vol. III. Morgan Kaufmann, Los Altos, CA, 1990.
Mitchell T. Generalization as Search. Artificial Intelligence, 18, 203–226, 1982.
Mitchell T., Keller R.M. & Kedar-Cabelli S. Explanation based generalization: A unifying view. Machine Learning, 1, 47–80, 1986.
Muggleton S. & Buntine W. Machine Invention of First-order Predicates by Inverting Resolution. In Proc. Fifth International Conference on Machine Learning, Ann Arbor, MI, pp.339–352, 1988.
Newell A., Simon H. Human Problem Solving. Prentice-Hall (Eds), Englewood Cliffs, NJ, 1972.
Pearl J. On the Connection between the Complexity and Credibility of Inferred Models. Int. J. of General Systems, 4, 255–264, 1978.
Plaisted D. Theorem Proving with Abstraction. Artificial Intelligence, 16, 47–108, 1981.
Reiter R. On Closed World Databases. In Gallaire H. and Minker J. (Eds) Logic and Databases, Plenum Press, New York, pp. 55–76, 1978.
Tenenberg J. Preserving Consistency across Abstraction Mappings. Proc IJCAI-87, pp. 1011–1014, Milano, Italy, 1987.
Utgoff P. Shift of Bias in Inductive Concept Learning. In R. S. Michalski, J. Carbonell, and T. Mitchell (Eds) Machine Learning, An Artificial Intelligence Approach, Vol. II, Morgan Kaufmann, Los Altos, CA, pp. 107–148, 1985.
Van de Velde W. Representation Issues in Second Generation Expert Systems. In K. Morik (Ed), Knowledge Representation and Organization in Machine Learning, Lecture Notes in Artificial Intelligence, Vol. 347, Springer-Verlag, pp. 17–49, 1989.
Watanabe S. Knowing and Guessing, A formal and Quantitative Study. Wiley Publ. Co, 1969.
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1991 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Giordana, A., Roverso, D., Saitta, L. (1991). Abstracting background knowledge for concept learning. In: Kodratoff, Y. (eds) Machine Learning — EWSL-91. EWSL 1991. Lecture Notes in Computer Science, vol 482. Springer, Berlin, Heidelberg. https://doi.org/10.1007/BFb0017000
Download citation
DOI: https://doi.org/10.1007/BFb0017000
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-53816-5
Online ISBN: 978-3-540-46308-5
eBook Packages: Springer Book Archive