Abstract
General-purpose generative planners use domain-independent search heuristics to generate solutions for problems in a variety of domains. However, in some situations these heuristics force the planner to perform inefficiently or obtain solutions of poor quality. Learning from experience can help to identify the particular situations for which the domain-independent heuristics need to be overridden. Most of the past learning approaches are fully deductive and eagerly acquire correct control knowledge from a necessarily complete domain theory and a few examples to focus their scope. These learning strategies are hard to generalize in the case of nonlinear planning, where it is difficult to capture correct explanations of the interactions among goals, multiple planning operator choices, and situational data. In this article, we present a lazy learning method that combines a deductive and an inductive strategy to efficiently learn control knowledge incrementally with experience. We present hamlet, a system we developed that learns control knowledge to improve both search efficiency and the quality of the solutions generated by a nonlinear planner, namely prodigy4.0. We have identified three lazy aspects of our approach from which we believe hamlet greatly benefits: lazy explanation of successes, incremental refinement of acquired knowledge, and lazy learning to override only the default behavior of the problem solver. We show empirical results that support the effectiveness of this overall lazy learning approach, in terms of improving the efficiency of the problem solver and the quality of the solutions produced.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
Aha, D. W., Kibler, D. & Albert, M. K. (1991). Instance-based learning algorithms. Machine Learning 6(1): 37–66.
Bhatnagar, N. (1992). Learning by incomplete explanations of failures in recursive domains. In Proceedings of the Ninth International Conference on Machine Learning, pp. 30–36, Aberdeen, Scotland: Morgan Kaufmann.
Borrajo, D., Caraça-Valente, J. P. & Morant, J. L. (1992a). Learning heuristics in planning. In Proceedings of the Sixth International Conference on Systems Research, Informatics and Cybernetics, pp. 43–49, Baden-Baden, Germany: The International Institute for Advanced Studies in Systems Research and Cybernetics.
Borrajo, D., Caraça-Valente, J. P. & Pazos, J. (1992b). A knowledge compilation model for learning heuristics. In Proceedings of the First Workshop on Knowledge Compilation, Aberdeen, Scotland.
Borrajo, D. & Veloso, M. (1994). Incremental learning of control knowledge for nonlinear problem solving. In Proceedings of the European Conference on Machine Learning, pp. 64–82. Catania, Italy: Springer Verlag.
Carbonell, J. G., Blythe, J., Etzioni, O., Gil, Y., Joseph, R., Kahn, D., Knoblock, C., Minton, S., Pérez, A., Reilly, S., Veloso, M. & Wang, X. (1992). PRODIGY4.0: The manual and tutorial. Technical Report CMU-CS–92–150, SCS, Carnegie Mellon University.
Carbonell, J. G., Knoblock, C. A. & Minton, S. (1990). Prodigy: An integrated architecture for planning and learning. In VanLehn, K. (ed.), Architectures for Intelligence, Erlbaum, Hillsdale, NJ. Also Technical Report CMU-CS–89–189.
Clark, P. & Holte, R. (1992). Lazy partial evaluation: An integration of explanation-based generalisation and partial evaluation. In Proceedings of the Ninth International Conference on Machine Learning, pp. 82–91, Aberdeen, Scotland: Morgan Kaufmann.
Cohen, W. W. (1990). Learning approximate control rules of high utility. In Proceedings of the Seventh International Conference on Machine Learning, pp. 268–276, Austin, TX: Morgan Kaufmann.
DeJong, G. F. & Mooney, R. (1986). Explanation-based learning: An alternative view. Machine Learning 1(2): 145–176.
Doorenbos, R. B. & Veloso, M. M. (1993). Knowledge organization and the utility problem. In Proceedings of the Third International Workshop on Knowledge Compilation and Speedup Learning, pp. 28–34, Amherst, MA.
Estlin, T. A. & Mooney, R. (1995). Hybrid learning of search control for partial order planning. In New Directions in AI Planning, pp. 115–128. IOS Press.
Etzioni, O. (1993). Acquiring search-control knowledge via static analysis. Artificial Intelligence 62(2): 255–301.
Etzioni, O. & Minton, S. (1992). Why EBL produces overly-specific knowledge: A critique of the Prodigy approaches. In Proceedings of the Ninth International Conference on Machine Learning, pp. 137–143. Aberdeen, Scotland. Morgan Kaufmann.
Fikes, R. E., Hart, P. E. & Nilsson, N. J. (1972). Learning and executing generalized robot plans. Artificial Intelligence 3: 251–288.
Hammond, K. J. (1989). Case-based Planning: Viewing Planning as a Memory Task. New York, NY: Academic Press.
Hanks, S. & Weld, D. (1995). A domain-independent algorithm for plan adaptation. Journal of Artificial Intelligence Research 2: 319–360.
Kambhampati, S. (1989). Flexible Reuse and Modification in Hierarchical Planning: A Validation Structure Based Approach. PhD thesis, Computer Vision Laboratory, Center for Automation Research, College Park, MD: University of Maryland.
Kambhampati, S. & Kedar, S. (1991). Explanation based generalization of partially ordered plans. In Proceedings of the Ninth National Conference on Artificial Intelligence, pp. 679–685. Anaheim, CA: AAAI Press.
Katukam, S. & Kambhampati, S. (1994). Learning explanation-based search control rules for partial order planning. In Proceedings of the Twelfth National Conference on Artificial Intelligence, pp. 582–587. Seattle, WA: AAAI Press.
Kettler, B. P., Hendler, J. A., Andersen, A. W. & Evett, M. P. (1994). Massively parallel support for case-based planning. IEEE Expert 2: 8–14.
Laird, J. E., Rosenbloom, P. S. & Newell, A. (1986). Chunking in SOAR: The anatomy of a general learning mechanism. Machine Learning 1: 11–46.
Langley, P. (1983). Learning effective search heuristics. In Proceedings of the Eighth International Joint Conference on Artificial Intelligence, pp. 419–421, Los Altos, CA: Morgan Kaufmann.
Leckie, C. & Zukerman, I. (1991). Learning search control rules for planning: An inductive approach. In Proceedings of the Eighth International Workshop on Machine Learning, pp. 422–426, Evanston, IL: Morgan Kaufmann.
Minton, S. (1988). Learning Effective Search Control Knowledge: An Explanation-Based Approach. Boston, MA: Kluwer Academic Publishers.
Minton, S., Knoblock, C. A., Kuokka, D. R., Gil, Y., Joseph, R. L. & Carbonell, J. G. (1989). PRODIGY 2.0: The manual and tutorial. Technical Report CMU-CS–89–146, School of Computer Science, Carnegie Mellon University.
Mitchell, T. M., Keller, R. M. & Kedar-Cabelli, S. T. (1986). Explanation-based generalization: A unifying view. Machine Learning 1: 47–80.
Mitchell, T. M., Utgoff, P. E. & Banerji, R. B. (1983). Learning by experimentation: Acquiring and refining problem-solving heuristics. In R. S. Michalski, J. G. Carbonell & T. Mitchell (eds.), Machine Learning, An Artificial Intelligence Approach. Palo Alto, CA: Tioga Press.
Muggleton, S. (1992). Inductive Logic Programming. London: Academic Press Limited.
Pérez, M. A. & Carbonell, J. G. (1994). Control knowledge to improve plan quality. In Proceedings of the Second International Conference on AI Planning Systems, pp. 323–328, Chicago, IL: AAAI Press.
Pérez, M. A. & Etzioni, O. (1992). DYNAMIC: A new role for training problems in EBL. In Proceedings of the Ninth International Conference on Machine Learning, pp. 367–372, Aberdeen, Scotland: Morgan Kaufmann.
Porter, B. W., Bareiss, R. & Holte, R. (1990). Knowledge acquisition and heuristic classification in weak-theory domains. Artificial Intelligence 45: 229–263.
Quinlan, J. R. (1990). Learning logic definitions from relations. Machine Learning 5: 239–266.
Rich, E. (1983). Artificial Intelligence. McGraw-Hill, Inc.
Ruby, D. & Kibler, D. (1992). Learning episodes for optimization. In Proceedings of the Ninth International Conference on Machine Learning, pp. 379–384, Aberdeen, Scotland: Morgan Kaufmann.
Stanfill, C. & Waltz, D. (1986). Toward memory-based reasoning. Communications of the Association for Computing Machinery 29: 1213–1228.
Stone, P., Veloso, M. & Blythe, J. (1994). The need for different domain-independent heuristics. In Proceedings of the Second International Conference on AI Planning Systems, pp. 164–169, Chicago, IL: AAAI Press.
Tadepalli, P. (1989). Lazy explanation-based learning: A solution to the intractable theory problem. In Proceedings of the Eleventh International Joint Conference on Artificial Intelligence, pp. 694–700, San Mateo, CA: Morgan Kaufmann.
Valiant, L. (1984). A theory of the learnable. Communications of the ACM 27(11): 1134–1142.
Veloso, M. & Blythe, J. (1994). Linkability: Examining causal link commitments in partialorder planning. In Proceedings of the Second International Conference on AI Planning Systems, pp. 170–175, Chicago, IL: AAAI Press.
Veloso, M. & Borrajo, D. (1994). Learning strategy knowledge incrementally. In Proceedings of the Sixth IEEE International Conference on Tools with Artificial Intelligence, pp. 484–490, New Orleans, LO: IEEE Computer Society Press.
Veloso, M., Carbonell, J., Pérez, A., Borrajo, D., Fink, E. & Blythe, J. (1995). Integrating planning and learning: The PRODIGY architecture. Journal of Experimental and Theoretical AI 7: 81–120.
Veloso, M. M. (1989). Nonlinear problem solving using intelligent causal-commitment. Technical Report CMU-CS–89–210, School of Computer Science, Carnegie Mellon University.
Veloso, M. M. (1994a). Flexible strategy learning: Analogical replay of problem solving episodes. In Proceedings of the Twelfth National Conference on Artificial Intelligence, Seattle, WA: AAAI Press.
Veloso, M. M. (1994b). Planning and Learning by Analogical Reasoning. Springer Verlag.
Waldinger, R. (1981). Achieving several goals simultaneously. In Nilsson, N. J. & Webber, B. (eds.), Readings in Artificial Intelligence, pp. 250–271. Los Altos, CA: Morgan Kaufmann.
Zelle, J. & Mooney, R. (1993). Combining FOIL and EBG to speed-up logic programs. In Proceedings of the Thirteenth International Joint Conference on Artificial Intelligence, pp. 1106–1113, Chambery, France: Morgan Kaufmann.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Borrajo, D., Veloso, M. Lazy Incremental Learning of Control Knowledge for Efficiently Obtaining Quality Plans. Artificial Intelligence Review 11, 371–405 (1997). https://doi.org/10.1023/A:1006549800144
Issue Date:
DOI: https://doi.org/10.1023/A:1006549800144