基于流形学习的基因微阵列数据分类方法

A Classification Method Based on Manifold Learning for Gene Microarray Data

提出了一种结合流形学习方法与分类算法的基因微阵列数据分类模型,先用流形学习算法对基因微阵列数据进行降维处理,然后再对降维后的数据进行分类.在实验中将流形学习算法LLE、ISO-MAP、LE和LTSA与三种分类算法相结合,并与直接用高维数据进行分类的结果进行了比较,实验结果表明所提出的模型极大地提高了分类精度,同时也提高了分类算法的执行效率.

Each sample in gene microarray data contains thousands or even tens of thousands of genes. It is necessary to reduce the dimension of the data before classifying them for obtaining better classified results. Manifold learning, as a nonlinear dimension reduction method, can discover the intrinsic laws hidden in the high dimensional data and has been widely applied in areas such as pattern recognition. A model combining manifold learning with classified algorithms was proposed to classify microarray data. In the model, the dimen- sion of microarray data was firstly reduced with some manifold learning method. Then the data reduced the di- mension were classified. In experiments, several manifold learning algorithms including LLE, ISOMAP, LE and LTSA are combined with three classified methods. And the results are compared with those from directly