A method of dual-process sample selection for feature selection on gene expression data

International Journal of Physical Sciences

Quanjin Liu¹ 

关键词: Feature selection;    support vector machine;    fuzzy interactive self-organizing data algorithm (ISODATA);    dual-process sample selection.;   

DOI  :  10.5897/IJPS12.327

学科分类：物理（综合）

来源: Academic Journals

【 摘 要 】

A method of dual-process sample selection based on support vector machine (SVM) is proposed to select informative features in this paper. Samples in a training set are used to train a SVM model, and the samples excluding support vector samples are chosen to select critical features in the procedure of recursive feature elimination (RFE). The effect of the dual-process sample selection method on feature selection is evaluated using the classification and the clustering performance of the selected features. The proposed dual-process sample selection method is applied to five gene expression datasets, and the experimental results show that the method is useful to improve the performance of the feature selection method based on fuzzy interactive self-organizing data algorithm (ISODATA). This indicates the method is reliable and effective for selecting informative genes from gene expression data.

【 授权许可】

CC BY