Advertisement

Machine Learning

, Volume 5, Issue 1, pp 39–70 | Cite as

Acquiring Recursive and Iterative Concepts with Explanation-Based Learning

  • Jude W. Shavlik
Article

Abstract

In explanation-based learning, a specific problem's solution is generalized into a form that can be later used to solve conceptually similar problems. Most research in explanation-based learning involves relaxing constraints on the variables in the explanation of a specific example, rather than generalizing the graphical structure of the explanation itself. However, this precludes the acquisition of concepts where an iterative or recursive process is implicitly represented in the explanation by a fixed number of applications. This paper presents an algorithm that generalizes explanation structures and reports empirical results that demonstrate the value of acquiring recursive and iterative concepts. The BAGGER2 algorithm learns recursive and iterative concepts, integrates results from multiple examples, and extracts useful subconcepts during generalization. On problems where learning a recursive rule is not appropriate, the system produces the same result as standard explanation-based methods. Applying the learned recursive rules only requires a minor extension to a PROLOG-like problem solver, namely, the ability to explicitly call a specific rule. Empirical studies demonstrate that generalizing the structure of explanations helps avoid the recently reported negative effects of learning.

Explanation-based generalization generalizing explanation structures generalizing to N generalizing number utility of learning operationality versus generality 

References

  1. Ahn, W., Mooney, R.J., Brewer, W.F., & DeJong, G.E (1987). Schema acquisition from one example: Psychological evidence for explanation-based learning. Proceedings of the Ninth Annual Conference of the Cognitive Science Society (pp.50-57). Seattle, WA: Lawrence Erlbaum.Google Scholar
  2. Anderson, J.R. (1986). Knowledge compilation: The general learning mechanism. In R.S. Michalski, J.G. Carbonell, & T.M. Mitchells (Eds.), Machine learning: An artificial intelligence approach (Vol.2). San Mateo, CA: Morgan Kaufmann.Google Scholar
  3. Andreae, P.M. (1984). Justified generalization: Acquiring procedures from example. Doctoral dissertation, Department of Electrical Engineering and Computer Science, Massachusetts Institute of Technology, Cambridge, MA.Google Scholar
  4. Bauer, M. (1979). Programming by examples. Artificial Intelligence, 12, 1-21.Google Scholar
  5. Biermann, A.W. (1978) The inference of regular LISP programs from examples. IEEE Transactions on Systems, Man, and Cybernetics, 8, 585-600.Google Scholar
  6. Cheng, P., & Carbonell, J.G. (1986). The FERMI system: Inducing iterative macro-operators from experience. Proceedings of the Fifth National Conference on Artificial Intelligence (pp.490-495). Philadelphia, PA: Morgan Kaufmann.Google Scholar
  7. Chien, S.A. (1989). Using and refining simplifications: Explanation-based learning of plans in intractable domains. Proceedings of the Eleventh International Joint Conference on Artificial Intelligence (pp.590-595). Detroit, MI: Morgan Kaufmann.Google Scholar
  8. Clocksin, W.F., & Mellish, C.S. (1984). Programming in PROLOG. Berlin: Springer Verlag.Google Scholar
  9. Cohen, W.W. (1987). A technique for generalizing number in explanation-based learning. Proceedings of the Fifth International Conference on Machine Learning (pp.256-269). Ann Arbor, MI: Morgan Kaufmann.Google Scholar
  10. Cohen, W.W. (1989). Solution path caching mechanisms which provably improve performance (Technical Report DCS-TR-254). New Brunswick, NJ: Rutgers University, Department of Computer Science.Google Scholar
  11. DeJong, G.F., & Mooney, R.J. (1986). Explanation-based learning: An alternative view. Machine Learning, I, 145-176.Google Scholar
  12. Dietterich, T.G., & Michalski, R.S. (1984). Discovering patterns in sequences of objects. Artificial Intelligence, 25, 257-294.Google Scholar
  13. Ellman, T. (1985). Generalizing logic circuit designs by analyzing proofs of correctness. Proceedings of the Ninth International Joint Conference on Artificial Intelligence (pp.643-646). Los Angeles, CA: Morgan Kaufmann.Google Scholar
  14. Fikes, R.E., Hart, P.E., & Nilsson, N.J. (1972). Learning and executing generalized robot plans. Artificial Intelligence, 3, 251-288.Google Scholar
  15. Green, C.C. (1969). Application of theorem proving to problem solving. Proceedings of the First International Joint Conference on Artificial Intelligence (pp.219-239). Washington, D.C.: Morgan Kaufmann.Google Scholar
  16. Hirsh, H. (1987). Explanation-based generalization in a logic-programming environment. Proceedings of the Tenth International Joint Conference on Artificial Intelligence (pp.221-227). Milan, Italy: Morgan Kaufmann.Google Scholar
  17. Jacobson, N. (1951). Lectures in abstract algebra (Vol.1). Princeton, NJ: Von Nostrand.Google Scholar
  18. Kedar-Cabelli, ST. (1986). Purpose-directed analogy: A summary of current research. In T.M. Mitchell, J.G. Carbonell, & R.S. Michalski (Eds.), Machine learning: A guide to current research. Hingham, MA: Kluwer.Google Scholar
  19. Kedar-Cabelli, S.T., & McCarty, L.T. (1987). Explanation-based generalization as resolution theorem proving. Proceedings of the Fourth International Workshop on Machine Learning (pp.383-389). Irvine, CA: Morgan Kaufmann.Google Scholar
  20. Keller, R.M. (1987). Defining operationality for explanation-based learning. Proceedings of the National Con-ference on Artificial Intelligence (pp.482-487). Seattle, WA: Morgan Kaufmann.Google Scholar
  21. Kodratoff, Y. (1979). A class of functions synthesized from a finite number of examples and a LISP program scheme. International Journal of Computer and Information Sciences, 8, 489-521.Google Scholar
  22. Korf, R.E. (1985). Depth-first iterative-deepening: An optimal admissible tree search. Artificial Intelligence, 27, 97-109.Google Scholar
  23. Laird, J.E., Rosenbloom, P.S., & Newell, A. (1986). Chunking in SOAR: The anatomy of a general learning mechanism. Machine Learning, 1, 11-46.Google Scholar
  24. Manna, Z. (1974). Mathematical theory of computation. New York: McGraw-Hill.Google Scholar
  25. McCarthy, J. (1963). Situations, actions, and causal laws (Memorandum). Stanford, CA: Stanford University, Department of Computer Science. (Reprinted in M. Minsky, (Ed.). Semantic information processing, 1968, Cambridge, MA: MIT Press.).Google Scholar
  26. Minton, S.N. (1988). Quantitative results concerning the utility of explanation-based learning. Proceedings of the Seventh National Conference on Artificial Intelligence (pp.564-569). St. Paul, MN: Morgan Kauftnann.Google Scholar
  27. Minton, S.N. (1989). Learning effective search control knowledge: An explanation-based approach. Hingham, MA: Kluwer.Google Scholar
  28. Mitchell, T.M., Keller, R.M., & Kedar-Cabelli, S. (1986). Explanation-based generalization: A unifying view. Machine Learning, 1, 47-80.Google Scholar
  29. Mitchell, T.M., Mahadevan, S., & Steinberg, L.I. (1985). LEAP: A learning apprentice for VLSI design. Pro-ceedings of the Ninth International Joint Conference on Artificial Intelligence (pp.573-580). Los Angeles, CA: Morgan Kaufmann.Google Scholar
  30. Mooney, R.J. (1988). Generalizing the order of operators in macro-operators. Proceedings of the Fifth International Conference on Machine Learning (pp.270-283). Ann Arbor, MI: Morgan Kaufmann.Google Scholar
  31. Mooney, R.J. (1989). The effect of rule use on the utility of explanation-based learning. Proceedings of the Eleventh International Joint Conference on Artificial Intelligence (pp.725-730). Detroit, MI: Morgan Kaufmann.Google Scholar
  32. Mooney, R.J. (in press). A general explanation-based learning mechanism and its application to narrative under-standing. London: Pitman.Google Scholar
  33. Mooney, R.J., & Bennett, S.W. (1986). A domain independent explanation-based generalizer. Proceedings of the Fifth National Conference on Artificial Intelligence (pp.551-555). Philadelphia, PA: Morgan Kaufmann.Google Scholar
  34. Prieditis, A.E. (1986). Discovery of algorithms from weak methods. Proceedings of the International Meeting on Advances in Learning (pp.37-52). Les Arcs, Switzerland.Google Scholar
  35. Rajamoney, S., & DeJong, G.F. (1987). The classification, detection, and handling of imperfect theory problems. Proceedings of the Tenth International Joint Conference on Artificial Intelligence (pp.205-207). Milan, Italy: Morgan Kaufmann.Google Scholar
  36. Riddle, P.J. (1989). Automating shifts of problem representation. Doctoral dissertation, Department of Computer Science, Rutgers University, New Brunswick, NJ.Google Scholar
  37. Sammut, C.B. (1986). Learning concepts by asking questions. In R.S. Michalski, J.G. Carbonell, & T.M. Mitchell (Eds.), Machine learning: An artificial intelligence approach (Vol.2). San Mateo, CA: Morgan Kaufmann.Google Scholar
  38. Segre, A.M. (1987). On the operationality/generality trade-off in explanation-based learning. Proceedings of the Tenth International Joint Conference on Artificial Intelligence (pp.242-248). Milan, Italy: Morgan Kaufmann.Google Scholar
  39. Shavlik, J.W. (in press). Extending explanation-based learning by generalizing the structure of explanations. London: Pitman.Google Scholar
  40. Shavlik, J.W., & DeJong, G.F. (1985). Building a computer model of learning classical mechanics. Proceedings of the Seventh Annual Conference of the Cognitive Science Society (pp.351-355). Irvine, CA: Morgan Kaufmann.Google Scholar
  41. Shavlik, J.W., & DeJong, G.F. (1987). BAGGER: An EBL system that extends and generalizes explanations. Proceedings of the Sixth National Conference on Artificial Language (pp.516-520). Seattle, WA: Morgan Kaufmann.Google Scholar
  42. Shavlik, J.W., & DeJong, G.F. (in press). Learning in mathematically-based domains: Understanding and generalizing obstacle cancellations. Artificial Intelligence.Google Scholar
  43. Shavlik, J.W., DeJong, G.F., & Ross, B.H. (1987). Acquiring special case schemata in explanation-based learn-ing. Proceedings of the Ninth Annual Conference of the Cognitive Science Society (pp.351-355). Irvine, CA: Morgan Kaufmann.Google Scholar
  44. Shavlik, J.W., & Maclin, R. (1988). An approach to acquiring algorithms by observing expert behavior. Pro-ceedings of the AAAI-88 Workshop on Automating Software Design (pp.172-181). St. Paul, MN.Google Scholar
  45. Shell, P., & Carbonell, J.G. (1989). Towards a general framework for composing disjunctive and iterative macro-operators. Proceedings of the Eleventh International Joint Conference on Artificial Intelligence (pp.596-602). Detroit, MI: Morgan Kaufmann.Google Scholar
  46. Smith, D.E., & Genesereth, M.R. (1985). Ordering conjunctive queries. Artificial Intelligence, 26, 171-215.Google Scholar
  47. Smith, D.R. (1984). The synthesis of LISP programs from examples: A survey. In A. Biermann, G. Guiho, and Y. Kodratoff, (Eds.), Automatic program construction techniques. New York: MacMillan.Google Scholar
  48. Summers, P.D. (1977). A methodology for LISP program construction from examples. Journal of the Association for Computing Machinery, 24, 161-175.Google Scholar
  49. Weld, D.S. (1986). The use of aggregation in casual simulation. Artificial Intelligence, 30, 1-34.Google Scholar

Copyright information

© Kluwer Academic Publishers 1990

Authors and Affiliations

  • Jude W. Shavlik
    • 1
  1. 1.Computer Sciences DepartmentUniversity of WisconsinMadison

Personalised recommendations