Data Reduction for Instance-Based Learning Using Entropy-Based Partitioning
Instance-based learning methods such as the nearest neighbor classifier have proven to perform well in pattern classification in several fields. Despite their high classification accuracy, they suffer from a high storage requirement, computational cost, and sensitivity to noise. In this paper, we present a data reduction method for instance-based learning, based on entropy-based partitioning and representative instances. Experimental results show that the new algorithm achieves a high data reduction rate as well as classification accuracy.
KeywordsData Reduction Irrelevant Attribute Euclidean Distance Measure Data Reduction Method Representative Instance
Unable to display preview. Download preview PDF.
- 2.Cano, J.R., Herrera, F., Lozano, M.: On the combination of evolutionary algorithms and strafitied strategies for training set selection in data mining. Applied Soft Computing (2005) (In Press, Correted Proof)Google Scholar
- 3.Datta, P., Kibler, D.: Learning prototypical concept description. In: Proceedings of the 12th International Conference on Machine Learning, pp. 158–166 (1995)Google Scholar
- 4.Datta, P., Kibler, D.: Symbolic nearest mean classifier. In: Proceedings of the 14th National Conference of Artificial Intelligence, pp. 82–87 (1997)Google Scholar
- 7.Dasarath, B.V.: Nearest Neighbor Norms: NN Pattern Classification Techniques. IEEE Computer Society Press, Los Alamitos (1991)Google Scholar
- 10.Han, J., Kamber, M.: Data Mining: Concepts and Techniques. Morgan Kaufmann, San Francisco (2001)Google Scholar
- 11.Merz, C.J., Murphy, P.M.: UCI Repository of Machine Learning Databases, Internet http://www.ics.uci.edu/~mlearn/MLRepository.html