基于PCA与核PCA的微阵列数据分析  

Microarray Data Analysis Based on PCA and Kernel PCA

作　　者：黄紫成 林增坦 HUANG Zi-cheng;LIN Zeng-tan(College of Engineering Technology,Yang-En University,Quanzhou Fujian 362014,China)

机构地区：[1]仰恩大学工程技术学院

出　　处：《长春师范大学学报》2019年第6期48-51,共4页Journal of Changchun Normal University

基　　金：福建省中青年教师教育科研项目“基于特征选择算法的微阵列数据挖掘”(JT180671);仰恩大学校级科研培育计划项目“基于特征选择算法的微阵列数据挖掘”(KJ20183002)

摘　　要：微阵列是近年来发展的生物技术,其产生数据典型特征是维数高而样本少,且数据常常存在缺少或噪声问题。在分析数据时,采用计算t值预处理方法解决此问题,同时针对数据高维的特点,采用PCA与核PCA对数据进行特征处理,然后应用支持向量机(SVM)进行训练,计算分类识别率。实验结果表明,经过降维处理之后能得到更高的分类识别率,提高了微阵列数据分析的准确性。Microarrays are biotechnologies that have been developed in recent years. The typical characteristics of their data generation are high dimensionality and few samples, and data often exists the lacks or noise problems. In the analysis of data, the pre-processing method of calculating t-value is used to solve this problem. At the same time, according to the characteristics of high-dimensional data, PCA and Kernel PCA are used to characterize the data, and then support vector machine is used for training and calculation,and calculativing classification recognition rate. The experimental results show that after the dimension reduction process, the higher classification recognition rate can be obtained, and the accuracy of microarray data analysis is improved.

关 键 词：微阵列 PCA 核PCA 支持向量机 

分 类 号：TP3-05[自动化与计算机技术—计算机科学与技术]