Problem decomposition and the learning of skills

Invited Papers
Part of the Lecture Notes in Computer Science book series (LNCS, volume 912)


One dimension of “divide and conquer” in problem solving concerns the domain and its subdomains. Humans learn the general structure of a domain while solving particular learning problems in it. Another dimension concerns the solver's goals and subgoals. Finding good decompositions is a major AI tactic both for defusing the combinatorial explosion and for ensuring a transparent end-product. In machine learning, pre-occupation with free-standing performance has led to comparative neglect of this resource, illustrated under the following headings. 1. Automatic manufacture of new attributes from primitives (“constructive induction”). 2. Machine learning within goal-subgoal hierarchies (“structured induction”). 3. Reconstruction of skills from human performance data (“behavioural cloning”).


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Muggleton, S. and Feng, C.: Efficient induction of logic programs. In Proc. First Conf. on Algorithmic Learning Theory (ed. S. Arikawa, S. Goto, S. Ohsuga and T. Yokomori, 1990), Tokyo: Japan. Soc. Art. Intell. pp. 368–381.Google Scholar
  2. 2.
    Sternberg, M.J.E., King, R.D., Lewis, R.A. and Muggleton,: S. Application of machine learning to structural molecular biology. Phil. Trans. R. Soc. Lond. B, 344 (1994) 365–371.Google Scholar
  3. 3.
    Newell, A.: A Unified Theory of Cognition. Boston, MA: Harvard University Press (1990).Google Scholar
  4. 4.
    Muggleton, S.H., Bain, M., Michie, D. and J. E. Hayes Michie.: An experimental comparison of human and machine learning formalisms. In Proc. Sixth Internat. Workshop on Machine Learning, (ed. A. Segre), Morgan Kaufman (1989) pp. 113–118.Google Scholar
  5. 5.
    Michie, D. and Bain, M. Machine acquisition of concepts from sample data.: In Artificial Intelligence and Intelligent Tutoring Systems, (eds. D. Kopec and R.B. Thompson), Ellis Horwood (1992) pp. 5–23.Google Scholar
  6. 6.
    Michalski, R.S. and Larson, J.B.: Inductive inference of VL decision rules. SIGART Newsletter, ACM, 63 (1977) 38–44.Google Scholar
  7. 7.
    Michie, D., Muggleton, S., Page, D. and Srinivasan., A.: To the International Computing Community: a New East-West Challenge (1994). Available as a Postscript file in URL See also URL Scholar
  8. 8.
    Hayes Michie, J.E.: Comparisons of Human and Machine Assessments of Complexity. Unpublished typescript, available from Ms Hayes Michie, 73860 Shadow Mountain Drive, Apt 2, Palm Desert, CA 92260, USA.Google Scholar
  9. 9.
    Turney, P.: Low size-complexity Inductive Logic Programming: the East-West clallenge considered as a problem in cost-sensitive classification. Submitted to IJCAI-95 (1995).Google Scholar
  10. 10.
    Turney, P.: Cost-sensitive classification:empirical evaluation of a hybrid decision tree induction algorithm. Jour. AI. Research. In press.Google Scholar
  11. 11.
    Lavrac, N. and Dzeroski, S.: Inductive Logic Programming: Techniques and Applications. New York: Ellis Horwood (1994).Google Scholar
  12. 12.
    Grefestette, J. J.: Optimization of control parameters for genetic algorithms. IEEE Trans. on Systems, Man and Cybernetics 16 (1986) 122–128Google Scholar
  13. 13.
    Shannon, C.E.: Computers and automata. Proc. I.R.E, 41 No. 10 (1953).Google Scholar
  14. 14.
    Shapiro, A.D.: Structured Induction in Expert Systems. Addison Wesley (1987).Google Scholar
  15. 15.
    Gaines, B.R.: Inducing Knowledge. Unpublished typescript. Available from Prof. Gaines at Knowledge Science Institute, University of Calgary, Alberta, Canada T2N 1N4. Email: (1995).Google Scholar
  16. 16.
    Sterling, L. and Shapiro, E.: The Art of Prolog. Cambridge, MA: The MIT Press (1986).Google Scholar
  17. 17.
    Michie, D.: New commercial opportunities using information technology. In Knowledge-based Systems: Third Internat Congr (1989).Google Scholar
  18. 18.
    Michie, D.: Problems of machine-aided concept formation. In Applications of Expert Systems, Vol 2 (ed. J.R. Quinlan), Addison-Wesley (1989) pp. 310–333.Google Scholar
  19. 19.
    Chambers, R. A. and Michie, D.: Man-machine co-operation on a learning task. In Computer Graphics — Techniques and Applications (eds. R. D. Parslow, R. W. Prowse and R. E. Green), Plenum Press (1969) pp. 179–185.Google Scholar
  20. 20.
    Urbancic, T. and Bratko, I.: Reconstructing human skill with machine learning. In Proc. Europ. Conf. on AI (ECAI 94), Amsterdam (1994).Google Scholar
  21. 21.
    Michie, D., Bain, M. and Michie, J.E.: Cognitive models from subcognitive skills. In Knowledge-Based Systems in Industrial Control (eds. M. Grimble, J. McGhee and P. Mowforth), Stevenage, Peter Peregrinus (1990).Google Scholar
  22. 22.
    Sammut, C., Hurst, S., Kedzier, D. and Michie, D.: In Proc. Ninth Internat. Workshop on Machine Learning, Morgan Kaufmann (1992) pp. 385–393.Google Scholar
  23. 23.
    Kibira, D.: Developing an expert controller of a black box simulation of a telephone line using machine induction. Unpublished report, Sydney: AI Lab, Univ. of New South Wales (1993).Google Scholar
  24. 24.
    Pearson, D. J., Huffman, S. B., Willis, M. B., Laird, J. E., and Jones, R. M.: A symbolic solution to intelligent real-time control. In Robotics and Autonomous Systems, 11, Elsevier (1993) pp. 279–291.Google Scholar
  25. 25.
    Benson, S. and Nilsson, N.: Reacting, planning, and learning in an autonomous agent. In Machine Intelligence 14 (eds. K. Furukawa, D. Michie and S. Muggleton), Oxford: the Clarendon Press, (in press).Google Scholar

Copyright information

© Springer-Verlag 1995

Authors and Affiliations

  1. 1.University of EdinburghUK
  2. 2.Palm DesertUSA

Personalised recommendations