Breast Cancer Prediction with High Accuracy using SVM: A Comparative Study of Machine Learning Model
Narendra Bavisetti, Saritha Abburi, J L Sarwani Theeparthi, Narasimhula Balayesu, Sudha Rani Mithikala, K. John Paul · 2025
Breast cancer has been a prevalent but most common disease among women due to the interplay of various clinical, lifestyle, societal, and economic factors. However, ML techniques have proven to be particularly effective at identifying such as through pattern recognition in datasets for breast cancer. This study is aimed at evaluating several machine learning algorithms for the prediction of breast cancer based on laboratory, demographic and mammography data. The analytical study utilized a dataset by Motamed Cancer Institute (ACECR) Tehran, Iran, with a total population of 5,178 records; of which around 25% were biopsy results of breast cancer patients. The dataset consisted of 24 variables for each record. A number of algorithms were employed such as logistic regression, k nearest neighbors (KNN), support vector machines (SVM), naive bayes classification, random forests, and decision trees. The Support Vector Machines (SVM) model was able to reach an impressive level of performance with an accuracy of 99.92%. Additionally, SVM model demonstrated the lowest false positive rate (FPR) of 0.02%, strong precision (98.23%), and strong recall (97.56%).