Logistic Regression for Breast Cancer Analysis
Bhoomi Sharma, Abhimanyu Abhimanyu, Anuradha Anuradha, Yogita Gigras · CiiT international journal of data mining and knowledge engineering · 2017
In this study, logistic regression on mammograms is used to diagnose breast cancer. The aim of using logistic regression is to obtain the significant clinical factors contributing more towards higher probability of breast cancer. The sample data set is taken from UC Irvine repository and modeled using the regression model. A 10-fold cross validation is applied on the training data set to avoid the over fitting problem. The sample data set contains mammograms samples collected by a survey conducted by the Radiologist. The classification table of 450 samples illustrations the correct classification percentage for mammogram as 96.6%. The result is then compared with 30 validated samples, correct classification 68.9%.The simulation results claims that the used linear regression model is able to map relationships among attributes by giving more accurate classification