Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
2576987 | International Congress Series | 2006 | 4 Pages |
Abstract
This paper proposes a hybrid real-coded genetic algorithm with forgetting for improving the generalization ability of classification models. A crucial idea here is the introduction of structural learning with forgetting into a hybrid real-coded genetic algorithm. The proposed method has two advantages: (1) finding near optimal classification models efficiently by a hybrid technique and (2) improving the generalization ability of the resulting classification models by the forgetting technique. Applications of the proposed method to an iris classification problem well demonstrate its effectiveness. Our results indicate that it has not only high learning performance for training data, but also high generalization ability for the test data compared with conventional algorithms such as backpropagation and structural learning with forgetting.
Related Topics
Life Sciences
Biochemistry, Genetics and Molecular Biology
Molecular Biology
Authors
H. Zhang, M. Ishikawa,