Analytical negative generalization and empirical negative generalization are not cumulative: A case study
This paper addresses the problem of using explanation-based generalizations of negative inputs in Mitchell's candidate elimination algorithm and in other more recent approaches to version spaces. It points out that a mere combination would produce a worse result than that obtainable without input preprocessing, whereas this problem was not perceived in previous work. Costs and benefits of the extra computation required to take advantage of prior EBL phase are analysed for a conjunctive concept languages defined on a tree-structured attribute-based instance space. Furthermore, this result seems to be independent of the particular inductive learning algorithm considered (i.e., version spaces), thus helping clarify one aspect of the ill-understood relation between analytical generalization and empirical generalization.
KeywordsInductive generalization analytical generalization version spaces
Unable to display preview. Download preview PDF.
- [Carp90]Carpineto, C. (1990). Combining EBL from success and EBL from failure with parameter version spaces. 9th European Conference on Artificial Intelligence. Stockholm.Google Scholar
- [Char86]Charniak, E., McDermott, D. (1986). Introduction to Artificial Intelligence. Addison-WesleyGoogle Scholar
- [Hirs89]Hirsh, H. (1989). Combining Empirical and Analytical Learning with Version Spaces. 6th International Workshop on Machine Learning. Syracuse, New York.Google Scholar
- [Hirs90]Hirsh, H. (1990). Incremental Version-Space Merging 7th International Conference on Machine Learning. Austin, Texas.Google Scholar
- [Mint87]Minton, S., Carbonell, J. G. (1987). Strategies for Learning Search Control Rules:An Explanation-based Approach. 10th IJCAI, Milan, Italy.Google Scholar
- [Mitc78]Mitchell, T. M. (1978). Version Spaces: An Approach to Concept learning. PhDthesis, Stanford University.Google Scholar
- [Mitc82b]Mitchell, T.M., Utgoff, P. E., Banerji, R.B. (1982). Learning problem-solving heuristics by experimentation. In R.S. Michalski et al. (Eds), Machine Learning. Tioga Press, Palo Alto, California.Google Scholar
- [Utgo86]Utgoff, P.E. (1986). Shift of Bias for Inductive Concept Learning. In R.S. Michalski et al. (Eds), Machine Learning.Vol.II Morgan Kaufmann, Los Altos, California.Google Scholar
- [VanL89]VanLehn, K. (1989). Efficient Specialization of Relational Concepts. Machine Learning 4 (1), 99–106.Google Scholar
- [Wins75]Winston, P. H. (1975). Learning structural descriptions from examples. In P.H. Winston (Ed.), The psychology of computer vision. New York: McGraw-Hill.Google Scholar