Machine Learning Approaches for Breast Cancer Prediction: A WEKA-Based Performance Analysis

The Major cause of deaths in women related to cancer is Breast Cancer these days. Increasing the accuracy in diagnostic is still a challenge even after so much progress in detecting and treating the disease. In this article, by using Machine Learning (ML) techniques, early investigation and early breast cancer prediction by focusing on two pivotal attributes: Menopause and Node-Caps. Utilizing Records of 1,144 patients each with ten attributes derived from testing reports, the study employs the Weka 3.6.9 tool in analyzing the dataset and compare the effectiveness of various ML classifiers, including KStar, Decision Table, and REPTree. For each classifier, the evaluation ensures calculation of F-Measure, Precision and Recall. The results indicate that KStar attains exorbitant accuracy, with weighted averages of 0.976 for Menopause and 0.980 for Node-Caps, indicating its worthier ability to forecast disease outcomes. Decision Table also exhibits a robust performance, especially in recall and F-Measure, while REPTree shows varying outcomes with improved performance for Node-Caps but lesser precision overall. The results emphasize the efficacy of ML techniques in augmenting early breast cancer detection, with KStar emerging as the most precise classifier for the traits analyzed. The study indicates that incorporating advanced ML methods could help in early diagnostics of breast cancer.