A Comparative Study of Classification Methods For Microarray Data Analysis

In response to the rapid development of DNA Microarray technology, many classification methods have been used for Microarray classification. SVMs, decision trees, Bagging, Boosting and Random Forest are commonly used methods. In this paper, we conduct experimental comparison of LibSVMs, C4.5, BaggingC4.5, AdaBoostingC4.5, and Random Forest on seven Microarray cancer data sets. The experimental results show that all ensemble methods outperform C4.5. The experimental results also show that all five methods benefit from data preprocessing, including gene selection and discretization, in classification accuracy. In addition to comparing the average accuracies of ten-fold cross validation tests on seven data sets, we use two statistical tests to validate findings. We observe that Wilcoxon signed rank test is better than sign test for such purpose.

• Type: Conference
• Year: 2006
• Where: AUSDM
• Authors: Hong Hu, Jiuyong Li, Ashley W. Plank, Hua Wang, Grant Daggard
• Keywords: AUSDM 2006, Data Mining, Data Sets, DNA Microarray Technology, Random Forest