流行学习算法应用于基因芯片数据分析

基因芯片是近年发展起来的生物技术,其数据典型特征是基因数多而样本少,因此必须及时采取有效方法来处理这些以指数级增长的数据.流行学习算法在高维数据方面有着广泛应用,但在基因芯片数据分析的应用还比较少.为了能得到在基因芯片数据分析中更好的处理方法,文章应用三种非线性降维提取海量基因芯片数据的特征,然后利用支持向量机作为分类器,判断样本的类属.实验结果表明,通过LLE特征提取之后,能获得与原基因芯片更为接近的成分,类属判断结果更为准确,为基因芯片数据分析提供了一定的科学指导.

Analysis on manifold learning algorithm applied to gene chip data

Gene chip is a recent development of biotechnology, where the characteristic of the data is that it has a large number of genes and few samples. Therefore, it is of great necessity for us to take effective measures to deal with the data which increases exponentially. Manifold learning is widely used in high dimensional data but less used in gene chip data. Three kinds of nonlinear dimension reduction methods are used to extract the feature of massive gene chip data. Then the support vector machine is utilized as base classifier to classify samples. The experimental results show that the composition can get closer to the original gene chip and classify more accurately after feature is extracted by LLE, and it provides some scientific guidance for the gene data analysis.