Learning in knowledge based systems, a possibilistic approach

  • Zbigniew W. Ras
  • Maria Zemankova
Late Arrivals
Part of the Lecture Notes in Computer Science book series (LNCS, volume 233)


Let us now make the final conclusions concerning the learning process by examples based on the concepts of a possibility distribution function and a rough set. As was shown the interpretation JS has the power of reducing the uncertainty whether an element belongs or does not belong to a concept to be learned. This is an improvement over the interpretation MS which can not provide any information for elements in the boundary. It can be observed that the computed function GS gets closer to the interpretation JS as the number of the learned concepts grows larger. The measure of learning defined by us reflects the effect of arranging the set of concepts to be learned into a particular sequence. The optimization method that we have outlined can be used to improve the learning process.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Aikins, J., "A theory and methodology of inductive learning", Artificial Intelligence, Vol.20, No.3, 1983Google Scholar
  2. 2.
    Buchanan, B. and Duda, R., "Principles of rule-based expert systems", Advances in Computers, Vol.22, 1983Google Scholar
  3. 3.
    Bobrow, D. and Collins, A., Representation and understanding, Academic Press, New York, 1975Google Scholar
  4. 4.
    Dietterich, T. and Michalski, R., "Inductive learning of structural descriptions", Artificial Intelligence, Vol.16, No.3, 1981Google Scholar
  5. 5.
    Michalski, R., "A theory and methodology of inductive learning", Artificial Intelligence, Vol.20, 1983, 111–16Google Scholar
  6. 6.
    Michalski, R. and Chilausky, R., "Learning by being told and learning from examples", Internat. J. Policy Anal. Inform. Systems, Vol.4, No.2, 1980, 125–160Google Scholar
  7. 7.
    Pawlak, Z., "Rough sets", Internat. J.Comp. and Info. Sci., Vol.11, 1982, 341–366Google Scholar
  8. 8.
    Pawlak, Z., "Information systems — theoretical foundations", Information Systems, Vol.6, No.3, 1981, 205–218Google Scholar
  9. 9.
    Ras, Z., "An algebraic approach to information retrieval systems" Internat. J. Comp. and Info. Sci., Vol.11, 1982, 275–293Google Scholar
  10. 10.
    Ras, Z., Zemankova, M., "Rough sets based learning systems", Proc. of the Conference on Computation Theory in Zaborow, Poland, December 1984, Lecture Notes in Computer Science, Springer Verlag, No. 208, 265–275Google Scholar
  11. 11.
    Winston, P., "Learning structural descriptions from examples", MAC TR-76, MIT, September 1970Google Scholar
  12. 12.
    Winston, P., "Learning by augmenting rules and accumulating censors", Report AIM-678, Artificial Intelligence Laboratory MIT, 1982Google Scholar
  13. 13.
    Zemankova-Leech, M. and Kandel, A., "Uncertainty propagation to expert systems", Approximate Reasoning in Expert Systems, eds. Gupta, M., Kandel, A., Bandler, W. and Kiszka, North Holland Publishers, 1985 /to appear/Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1986

Authors and Affiliations

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

Personalised recommendations