郑州大学学报（理学版）2017，Vol.49Issue(4)：28-33,6.DOI:10.13705/j.issn.1671-6841.2017096
基于PCA和多邻域粗糙集的肿瘤特征基因选择算法
Tumor Feature Gene Selection Method Based on PCA and Multiple Neighborhood Rough Set
摘要
Abstract
To solve the problems in higher time complexity and blurry description toward the gene expres-sion profile in the approximation calculation using the global neighborhood , an effective PNRS model was proposed based on principal component analysis (PCA) and neighborhood rough set (NRS).First of all, the low dimensional feature space was obtained by using PCA algorithm;then the multiple neighborhood rough set algorithm was adopted for feature gene selection , namely calculating neighborhood attribute val-ues through Euclidean distance , followed by approximation of neighborhood decision system .Finally, feature gene set was obtained by using the heuristic search method .The experimental results showed that the PNRS model achieved higher classification accuracy with respect to smaller gene subsets .The simula-tion results showed the validity of the proposed method .关键词
特征选择/主成分分析/多邻域粗糙集/欧氏距离Key words
feature selection/principal component analysis/multiple neighborhood rough set/Euclidean distance分类
信息技术与安全科学引用本文复制引用
徐久成,穆辉宇,冯森..基于PCA和多邻域粗糙集的肿瘤特征基因选择算法[J].郑州大学学报（理学版）,2017,49(4):28-33,6.基金项目
国家自然科学基金项目(61370169,61402153) （61370169,61402153）
河南省科技攻关重点项目(142102210056,162102210261) （142102210056,162102210261）
河南师范大学青年科学基金项目(2014QK28) （2014QK28）
河南省高等学校重点科研项目(16A520057). （16A520057）
