期刊名称:Journal of Emerging Trends in Computing and Information Sciences
电子版ISSN:2079-8407
出版年度:2012
卷号:3
期号:10
页码:1419-1426
出版社:ARPN Publishers
摘要:In this study, two models for constructing acute leukemia classifiers using the Signal-to-Noise Ratio (SNR) gene selection method in conjunction with the Bayesian Networks (BNs) have been proposed. In the first model, genes of the acute leukemia training dataset are ranked using the SNR method and then top ranked genes are selected and used to construct the acute leukemia BN classifier. In the second model, genes of the acute leukemia training dataset are clustered using thek-means clustering and then genes of each cluster are ranked using the SNR method after that top ranked genes from gene clusters are selected and used to construct the acute leukemia BN classifier. From the experimental evaluation, the results showed that the classification accuracies achieved by the acute leukemia classifiers constructed according to either of these two models are good compared with the classification accuracies achieved in other studies. The results also indicated that the second model is better than the first model.