Learning by Experimentation: Acquiring and Refining Problem-Solving Heuristics

  • Tom M. Mitchell
  • Paul E. Utgoff
  • Ranan Banerji
Part of the Symbolic Computation book series (SYMBOLIC)


This chapter concerns learning heuristic problem-solving strategies through experience. In particular, we focus on the issue of learning heuristics to guide a forward-search problem solver, and describe a computer program called lex, which acquires problem-solving heuristics in the domain of symbolic integration. lex acquires and modifies heuristics by iteratively applying the following process: (i) generate a practice problem; (ii) use available heuristics to solve this problem; (iii) analyze the search steps performed in obtaining the solution; and (iv) propose and refine new domain-specific heuristics to improve performance on subsequent problems. We describe the methods currently used by lex, analyze strengths and weaknesses of these methods, and discuss our current research toward more powerful approaches to learning heuristics.


Assure Dition 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. Anzai, Y. and Simon, H., “The theory of learning by doing,” Psychological Review, Vol. 36, No. 2, pp. 124–140, 1979.CrossRefGoogle Scholar
  2. Buchanan, B. G. and Mitchell, T. M., “Model-Directed Learning of Production Rules,” Pattern-Directed Inference Systems, Waterman, D. A. and Hayes-Roth, F. (Eds.), Academic Press, New York, 1978.Google Scholar
  3. Davis, R., “Applications of meta level knowledge to the construction and use of large knowledge bases,” Knowledge-based Systems in Artificial Intelligence, Davis, R. and Lenat, D. (Eds.), McGraw-Hill, New York, 1981.Google Scholar
  4. Fikes, R. E., Hart, P. E. and Nilsson, N. J., “Learning and executing generalized robot plans,” Artificial Intelligence, Vol. 3, pp. 251–288, 1972.CrossRefGoogle Scholar
  5. Hayes-Roth, F., Klahr, P. and Mostow, D. J. J., “Knowledge acquisition, knowledge programming, and knowledge refinement”, Technical Report R-2540-NSF, The Rand Corporation, Santa Monica, CA., May 1980.Google Scholar
  6. Iba, G. A., “Learning disjunctive concepts from examples,” Master’s thesis, M.I.T., Cambridge, Mass., 1979, (also AI memo 548).Google Scholar
  7. Mitchell, T. M., Version Spaces: An approach to concept learning,Ph.D. dissertation, Stanford University, December 1978, (also Stanford CS report STAN-CS-78–711, HPP-79–2).Google Scholar
  8. Mitchell, T. M., Utgoff, P. E., Nudel, B. and Banerji, R., “Learning problem-solving heuristics through practice,” Proceedings of the Seventh International Joint Conference on Artificial Intelligence, Vancouver, pp. 127–134, August 1981.Google Scholar
  9. Mitchell, T. M., “Generalization as Search,” Artificial Intelligence, Vol. 18, No. 2, pp. 203–226, March 1982.MathSciNetCrossRefGoogle Scholar
  10. Mitchell, T. M., “Toward Combining Empirical and Analytic Methods for Learning Heuristics,” Human and Artificial Intelligence, Elithorn, A. and Banerji, R. ( Eds. ), Erlbaum, 1982.Google Scholar
  11. Neves, D. M. M., “A computer program that learns algebraic procedures,” Proceedings of the 2nd Conference on Computational Studies of Intelligence, Toronto, 1978.Google Scholar
  12. Nilsson, N. Principles of Artificial Intelligence, Tioga, Palo Alto, 1980.MATHGoogle Scholar
  13. Politakis, P., Weiss, S. and Kulikowski, C., “Designing consistent knowledge bases for expert consultation systems”, Technical Report DCS-TR-100, Department of Computer Science, Rutgers University, 1979, (also 13th Annual Hawaii International Conference on System Sciences).Google Scholar
  14. Utgoff, P. E. and Mitchell, T. M., “Acquisition of Appropriate Bias for Inductive Concept Learning,” Proceedings of the 1982 National Conference on Artificial Intelligence,Pittsburgh, August 1982.Google Scholar
  15. Vere, S. A., “Inductive learning of relational productions,” Pattern-Directed Inference Systems, Waterman, D. A. and Hayes-Roth, F. (Eds.), Academic Press, New York, 1978.Google Scholar
  16. Waterman, D. A., “Generalization learning techniques for automating the learning of heuristics,” Artificial Intelligence, Vol. 1, No. 1 /2, pp. 121–170, 1970.MATHGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1983

Authors and Affiliations

  • Tom M. Mitchell
    • 1
  • Paul E. Utgoff
    • 1
  • Ranan Banerji
    • 2
  1. 1.Rutgers UniversityUSA
  2. 2.St. Joseph’s UniversityUSA

Personalised recommendations