基于INCA的肿瘤基因表达谱分类模型  

Classification model of gene expression profile based on INCA

作　　者：潘江山 陈晓云[1] 董红玉[1] 

机构地区：[1]福州大学数学与计算机科学学院,福建福州350116

出　　处：《福州大学学报（自然科学版）》2014年第4期639-645,共7页Journal of Fuzhou University(Natural Science Edition)

基　　金：福建省自然科学基金资助项目(2014J01009);福建省新世纪优秀人才项目(XSJRC2007-11)

摘　　要：针对NCA算法对初始值敏感的不足,提出一种改进的NCA算法(INCA).INCA对肿瘤基因表达谱进行奇异值分解,将标准化后的右奇异矩阵作为初始值,提取肿瘤基因表达谱中的分类信息.在4个标准肿瘤基因表达谱数据集上进行实验,以INCA作为特征提取方法,K-近邻、Parzen窗作为分类器进行分类检测.实验结果表明,与NCA及现有的分类模型相比,基于INCA的分类模型能够取得较高的分类准确率.In view of the shortcoming that NCA is sensitive to initial value, an improved method INCA is proposed in this work. Firstly, INCA compute the SVD of tumor gene expression profile, then chose the standardized right singular matrix as the initial value of NCA algorithm, extracting the classification information in the tumor gene expression profile. Use K - NN, Parzen as classifier to evaluate the classi- fied effect of INCA on four standard tumor gene expression profile datasets. The results show that the ac- curacy of classification model based on INCA is higher than both NCA and other classification models.

关 键 词：肿瘤基因表达谱 奇异值分解 分类 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]