SVM-Based Framework for Breast Cancer Detection

Breast cancer has been among the most prevalent cancers in the last five years. According to WHO in 2020, more than 2.35 million people were diagnosed with Breast Cancer and more than 690 thousand deaths globally. Both researchers and doctors are facing the challenges of fighting cancer. This research paper aims to use different Supervised machine-learning techniques namely KNN, SVM (Support Vector Machine) and Logistic Regression for breast cancer detection. Our main objective is to determine whether the patient is diagnosed with a malign or benign cancer type. The Machine Learning algorithms were applied to the obtained dataset and the algorithms were evaluated using Accuracy as the performance measure. The obtained accuracy scores after applying KNN, SVM and LR are 92.98%, 95.5% and 95.11% respectively. The highest accuracy has been achieved by applying SVM. Early detection and accurate prediction of Cancer are fundamental to identifying patients who could benefit from the treatment and this can help in reducing the mortality rate due to cancer. With the help of various ML techniques, we can detect breast cancer more efficiently and effectively.