Combining Rule-Based and Sample-Based Classifiers – Probabilistic Approach
The present paper is devoted to the pattern recognition methods for combining heterogeneous sets of learning data: set of training examples and the set of expert rules with unprecisely formulated weights understood as conditional probabilities. Adopting the probabilistic model two concepts of recognition learning are proposed. In the first approach two classifiers trained on homogeneous data set are generated and next their decisions are combined using local weighted voting combination rule. In the second method however, one set of data is transformed into the second one and next only one classifier trained on homogeneous set of data is used. Furthermore, the important problem of consistency of expert rules and the learning set is discussed and the method for checking it is proposed.
Unable to display preview. Download preview PDF.
- 2.Czabanski, R.: Self-generating fuzzy rules from numerical data. Techn. Report Silesian Technical Univ. Gliwice (PhD Thesis) (2002) (in Polish)Google Scholar
- 4.Dubois, D., Lang, J.: Possibilistic logic. In: Handbook of Logic in Artificial Intelligence and Logic Programming, pp. 439–513. Oxford Univ. Press, Oxford (1994)Google Scholar
- 8.Kittler, J., Duin, R., Matas, J.: On combining classifiers. IEEE Trans. on PAMI 20, 226–239 (1998)Google Scholar
- 10.Kurzynski, M., Sas, J., Blinowska, A.: Rule-Based Medical Decision-Making with Learning. In: Proc. 12th World IFAC Congress, Sydney, vol. 4, pp. 319–322 (1993)Google Scholar
- 11.Kurzynski, M., Wozniak, M.: Rule-Based Algorithms with Learning for Sequential Recognition Problem. In: Proc. 3rd Int. Conf. Fusion 2000, Paris, pp. 10–13 (2000)Google Scholar
- 15.Kuratowski, K., Mostowski, A.: Set Theory. Nort-Holland Publishing Co., Amsterdam (1986)Google Scholar
- 17.Woods, K., Kegelmeyer, W.: Combination of multiple classifiers using local accuracy estimates. IEEE Trans. on PAMI 19, 405–410 (1997)Google Scholar