Explanation-based generalization in game playing: Quantitative results

  • Stefan Schrödl
Relational Learning
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1398)


Game playing has attracted researchers in Artificial Intelligence ever since its beginnings. By comparison with human reasoning, learning by operationalization of general knowledge, as formalized by the Explanation-Based Generalization (EBG) paradigm, appears to be highly plausible in this domain. Nevertheless, none of the previously published approaches is (provably) sufficient for the target concept, and at the same time applicable to arbitrary game states.

We trace this paradox back to the lack of the expressive means of Negation as Failure in traditional EBG, and constructively support our claim by applying the respective extension proposed in [Schr96] to the chess endgame king-rook vs. king-knight.

Methodically, endgames are well-suited for quantitative evaluation and allow to obtain more rigorous results concerning the effects of learning than in other domains. This is due to the fact that the entire problem space is known (and can be generated) in advance.

We present the main results of a large-scale empirical study. The issues of training complexity, speedup for recognition and classification, as well as the question of optimal reasoning under time constaints are analyzed.


Explanation-Based Learning applications 


  1. [Cla78]
    Clark, K.L., Negation as Failure, in: Gallaire, H., and Minker, J. (eds.), Logic and Data Bases, Plenum Press, NY, (1978), pp 293–322Google Scholar
  2. [FiNi71]
    Fikes, R. E., and Nilsson, N. J., STRIPS: a New Approach to the Application of Theorem Proving to Problem Solving, Artificial Intelligence 2(3–4), pp 189–208Google Scholar
  3. [FlaDi86]
    Flann, N.S., and Dietterich, T.G., Selecting Appropriate Representations for Learning from Examples, AAAI-86, (1986), pp 460–466Google Scholar
  4. [FlaDi89]
    Flann, N.S., and Dietterich, T.G., A Study of Explanation-Based Methods for Inductive Learning, Machine Learning 4, Kluwer Academic Publishers, Boston, (1989), pp 187–226Google Scholar
  5. [GoKe87]
    Gorlick, M.M., and Kesselman, C.F., Timing Prolog Programs Without Clocks, Symp. on Logic Programming, IEEE Computer Society, (1987), pp 426–432Google Scholar
  6. [KCMcC87]
    Kedar-Cabelli, S.T., and McCarty, L.T., Explanation-Based Generalization as Theorem Proving, Proc. of the Fourth International Workshop on Machine Learning, Irvine, (1987)Google Scholar
  7. [Ke188]
    Keller, R. M., Defining Operationality for Explanation-Based Learning, Artificial Intelligence 35, (1988), pp 227–241CrossRefGoogle Scholar
  8. [Min84]
    Minton, S., Constraint-Based Generalization: Learning Game-Playing Plans from Single Examples, AAAI-84, (1984), pp 251–254Google Scholar
  9. [Mi90]
    Minton, S., Quantitative Results Concerning the Utility of Explanation-Based Learning, Artificial Intelligence 42, Elsevier Publishers, (1990), pp 363–391Google Scholar
  10. [MKKC86]
    Mitchell, T., Keller, R., and Kedar-Cabelli, S., Explanation-Based Generalization: A Unifying View, Machine Learning 1:1, (1986), pp 47–80Google Scholar
  11. [Pu94]
    Puget, J.-F., Explicit Representation of Concept Negation, Machine Learning 14, Kluwer Academic Publishers, Boston, (1994), pp 233–247Google Scholar
  12. [Qu83]
    Quinlan, J. R., Learning Efficient Classification Procedures and their Application to Chess End Games, in: Michalski, et al. (eds.), Machine Learning: An Artificial Intelligence Approach, San Mateo, CA, Morgan Kaufmann, (1983), pp 463–482Google Scholar
  13. [Schr96]
    Schrödl, S., Explanation-Based Generalization for Negation as Failure and Multiple Examples, ECAI-96, Budapest, Hungary, John Wiley & Sons, (1996), pp 448–452 (1986)Google Scholar
  14. [SiPu88]
    Siqueira, J. L., and Puget, J. F., Explanation-Based Generalization of Failures, Proc. of ECAI-88, Munich, (1988), pp 339–344Google Scholar
  15. [Ta89]
    Tadepalli, P., Lazy Explanation-Based Learning: A Solution to The Intractable Theory Problem, IJCAI-89, Detroit MI, (1989), 694–700Google Scholar
  16. [Wa75]
    Waldinger, R., Achieving Several Goals Simultaneously; in: Elcock, E. W., and Michie, D. (eds.), Machine Intelligence 8, Ellis Horwood, Chichester, England, pp 94-138Google Scholar
  17. [YSUB90]
    Yee, R.C., Saxena, S., Utgoff, P.E., Barto, A.G., Explaining Temporal Differences to Create Useful Concepts for Evaluating States, AAAI-90, (1990), pp 882–888Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1998

Authors and Affiliations

  • Stefan Schrödl
    • 1
  1. 1.Institut für InformatikAlbert-Ludwigs-UniversitätFreiburgGermany

Personalised recommendations