A Hybrid Gene Selection Method for Microarray Data Based on Geodesic Distance and Binary Particle Swarm Optimization

2nd International Symposium on Application of Materials Science and Energy Materials

材料科学;能源学

Xiong, Ying^1 ; Han, Fei^1

School of Computer Science and Communication Engineering, Jiangsu University, Zhenjiang, China^1

关键词: Binary particle swarm optimization;    Clustering methods;    Gene selection;    Geodesic distances;    High-dimensional;    Intrinsic geometry;    Microarray data;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/490/4/042014/pdf
DOI  :  10.1088/1757-899X/490/4/042014

学科分类：材料科学（综合）

来源: IOP

【 摘 要 】

To obtain the most predictive genes subsets without filtering out critical genes, a method for gene selection based on binary particle swarm optimization (BPSO) and geodesic distance is proposed in this paper. In this approach, to preserve the intrinsic geometry of high dimensional microarray data, geodesic distance is calculated as the measurement between genes for cluster, and by combining with clustering method, BPSO is used to perform gene selection to reduce redundancy. With experiments conducted on several public microarray data by ELM classifiers, the results confirm that it is efficient to use the proposed method for gene selection compared to the relevant gene selection method.