基于杂交式基因选择的核Fisher分类算法

Kernel Fisher Discriminant Algorithm Based on Hybrid Gene Selection

为提高核Fisher算法(KFDA)的分类性能,对KFDA算法增加了基因选择步骤.提出了新型的杂交式基因选择算法并用于KFDA分类研究,所提出的方法用于3个基因芯片的数据分类,得到较好的分类效果.

Gene selection step is added to improve the classification performance of kernel Fisher discriminant algorithm（KFDA）. Hybrid gene selection methods are suggested for classifying microarray data. The proposed approach is used to three public datasets and shows good prediction performance.