Prediction of Breast Cancer of Women Based on Support Vector Machines

In recent years, the incidence of female breast cancer in China has increased rapidly, and breast cancer has become the most seriously threaten of Chinese women. Therefore, it is of great significance to make accurate prediction and judgment female breast tumors being benign or malignant. This study is based on Support Vector Machines (SVM) to classify and predict the female breast tumors benign or malignant. First, breast cancer data has been dealing with, then will be classified. We adopt SVM algorithm and three models including SVC (kernel = 'linear'), SVC (kernel = 'RBF) and linear SVC () and Nu SVC (). The classify algorithms are linear, nonlinear, nuclear. We combine the three algorithms with the previous models to forecast breast tumor benign or malignant. The classifier model is evaluated by integrating the indexes such as precision rate, recall rate and running time, and then the super parameters of the classifier are analyzed and optimized. Through trial-and-error, it can be found that model SVC (kernel= 'RBF') has the best prediction effect. The optimized classifier model can be achieved 98.86% accuracy, 95.29% recall and 95.46% harmonic average (F1-score).