SVM for Breast Cancer Detection from mammogram Using MatLAB
The code solves the Lagrange dual problem of SVM by using quadprog to find alpha, support vectors and bias
4 features 800 instances Missing values available
F1-score, Precision and recall of the test set then calculated
RBF found to be the best Kernel of SVM to classify the dataset
optimal sigma for RBF is choosen to be the one who brings the larger variance between the transformed data (generally easier to classify)
main code is here