Risk Prediction Modeling for Breast Cancer using Supervised Machine Learning Approaches

Millions of people worldwide are affected by breast cancer, a complicated and possibly fatal disease. For patients to have better outcomes and have higher survival rates, early detection and precise diagnosis are essential. The importance of machine learning methods in breast cancer research has grown as a result of their capacity to manage complicated and varied medical data. This study compares the performance of machine learning models on the Wisconsin Breast Cancer dataset, including Logistic Regression (LR), Support Vector Machine (SVM), Artificial Neural Networks (ANN), and Random Forest (RF). Through the use of a variety of factors, including accuracy, sensitivity, specificity, precision, & F1-Score, the research seeks to assess the models' performance. The findings demonstrate that these machine learning models predict breast cancer very effectively, with accuracy rates ranging from 95.71% to 98.57%. The ANN model detected every instance of breast cancer with the best accuracy and sensitivity. The performance of the RF and Logistic Regression models was balanced, however the SVM model had a little lower specificity. The F1-Scores show that these models have a good balance between accuracy and recall. These results demonstrate the possibility of early identification and enhanced patient care offered by machine learning in the prediction of breast cancer. Overall, ML has enormous promise for improving the prognosis and detection of breast cancer, lowering the chance of misdiagnosis, and expediting the diagnostic procedure, which will eventually benefit patients and the healthcare industry.