Advertisement

Learning driven by the concepts structure

  • Zbigniew W. Ras
  • Maria Zemankova
Knowledge Acquisition And Machine Learning
Part of the Lecture Notes in Computer Science book series (LNCS, volume 313)

Abstract

Initially, concepts are described in terms of values of attributes. These descriptions are in a probabilistic DNF form. Assuming a growing language, concepts already taught can be used in describing new concepts. The order of teaching the concepts is the key to producing their optimal descriptions. By "optimal" we mean the minimum number of occurences of constants in descriptions. The problem of finding the minimal description for each concept is NP-complete, hence our proposed algorithm has to be heuristic. Our strategy is based on clustering terms in concept descriptions in order to replace them by shorter higher level terms. Concepts are taught with respect to the increasing order of disjuncts in their descriptions. The outcome of the algorithm are optimized concepts descriptions in terms of a growing language, and a concept hierarchy that can be used for further learning and as a tool for generalization/specialization/similarity based reasoning.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Bibliography

  1. [1]
    Fisher, D.H., "Conceptual Clustering, Learning from Examples, and Inference", Proc. of the 4th Int. Workshop on Machine Learning, Irvine, California, June 22–25, 1987, Morgan Kaufmann Publ., 38–49Google Scholar
  2. [2]
    Langley, P., Gennari, J.H., Iba, W., "Hill-climbing Theories of Learning", ibid., 312–323Google Scholar
  3. [3]
    Michalski, R.S., "A Geometrical Model for Synthesis of Interval Covers", Report No. 461, Dept. of Comp. Sci., Univ. of Illinois, Urbana, 1971Google Scholar
  4. [4]
    Michalski, R.S., Larson, J.B., "Selection of Most Representative Training Examples and Incremental Generation of VL1 Hypothesis: the Underlying Methodology and Description of Programs ESEL and AQ11", Report No. 867, Dept. of Comp. Sci., Univ. of Illinois, Urbana, 1978Google Scholar
  5. [5]
    Pawlak, Z., "Rough Sets", Int. J. Comp. and Info. Sci., Vol. 11, 1982, 341–366Google Scholar
  6. [6]
    Quinlan, J.R., "Decision Trees as Probabilistic Classifiers", Proc. of the 4th Int. Workshop on Machine Learning, Irvine, California, June 22–25, 1987, Morgan Kaufmann Publ., 31–37Google Scholar
  7. [7]
    Ras, Z.W., Zemankova, M., "Learning in Knowledge Based Systems, a Possibilistic Approach", Proc. of the MFCS in Bratislava, Czechoslovakia, August 25–29, 1986, Lecture Notes in Computer Science, Springer Verlag, No. 233, 630–638Google Scholar
  8. [8]
    Wong, J.H., Wong, S.K.M., "An Inductive Learning System — ISL", Methodologies for Intelligent Systems, Z.W. Ras, M. Zemankova, Eds., North Holland, 1987, 370–378Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1988

Authors and Affiliations

  • Zbigniew W. Ras
    • 1
  • Maria Zemankova
    • 2
  1. 1.Dept. of Comp. Sci.Univ. of North CarolinaCharlotte
  2. 2.Dept. of Comp. Sci.Univ. of TennesseeKnoxville

Personalised recommendations