SNP selection based on k-center algorithm

Single Nucleotide Polymorphism (SNP) data has been used for genetic science research with a high level of data importance. However, there are more redundant SNP data and noise from a large number of SNP data and hence, to select the essential SNP characteristics is crucial. In this paper, k-center is used for data dimensionality reduction, and symmetric uncertainty is introduced into the distance measurement of k-center algorithm to solve the linkage imbalance between SNP data. An improved k-center algorithm is proposed, it is k - MSU algorithm, in the hospital to provide the experimental results of clinical trial data show that k - MSU algorithm in SNP selection has higher classification accuracy and the better effect.