Screening of Pathological Gene in Breast Cancer Based on Logistic Regression

摘要

Breast cancer has become the focus of the pathological gene screening research. In this paper, logistic regression and multiple hypothesis testing are used to screen the pathological gene based on the existing breast cancer genetic data. Then referencing the confidence level, the p-value of logistic regression is used to screen the pathological gene initially. Furthermore, by considering the Type I error, the multiple hypothesis testing is used to make the result accurate. In addition, SVM is used to test the reliability of this paper's methods. In order to illustrate the feasibility of this method, each gene which screened by this method is tested and verified by the literature of breast cancer.