A Comparative Analysis of Machine Learning Algorithms for Breast Cancer Prediction and Detection

Abstract

Many efforts have been devoted to developing efficient methods for early detection and prediction in order to address this urgent problem. My suggested strategy makes use of risk prediction algorithms and encourages recent breast cancer awareness. My strategy offers a simple way to forecast breast cancer and stands out for its practical usefulness in real-world circumstances. Using dataset provided on the UCI platform, I included a variety of classifiers, such as “K-Nearest Classifier (KNN), Random Forest (RF), Decision Tree (DT), Support Vector Machine (SVM), Naïve Bayes, and Logistic Regression (LR)”. The findings were encouraging. Moreover, Boosting Decision Tree showed accuracy that met this exacting requirement. I applied a variety of techniques, including as Bagging, Boosting, and Voting, to further improve performance. In addition to adding to the corpus of information on breast cancer detection and prediction, my experimental analysis revealed the most accurate model to be the Logistic Regression model, which achieved an exceptional accuracy rate of 97.37% for breast cancer. The main aim of this thesis patient outcomes enhances by presenting a viable early intervention approach and important new insights into the management.