An Ensemble Method in Hybrid Real-Coded Genetic Algorithm with Pruning for Data Classification

H. Zhang and M. Ishikawa (Japan)


ensemble method, hybrid real-coded genetic algorithm, pruning, data classification


To obtain a classification model with high generaliza tion ability, this paper proposes a novel ensemble method that implements a hybrid real-coded genetic algorithm with pruning (HRGA/P). A crucial idea here is to combine en semble learning and HRGA/P with parallel computational ability and high generalization ability. Accordingly, the re sulting classification model is expected to have high gen eralization ability. Applications of the proposed method to a wine classification problem well demonstrate its effec tiveness. The characteristics of generalization ability of an interpolated model from two classification models are also investigated.

Important Links:

Go Back