A Machine Learning Approach to Predict Breast Cancer Using Boosting Classifiers

Abstract

Breast cancer is a prevalent disease, with the second highest incidence rate among all types of cancer. The risk of death from breast cancer is increasing due to rapid population growth, and a dependable and quick diagnostic system can assist medical professionals in disease diagnosis and lower the mortality rate. In this study, various machine-learning algorithms are examined for predicting the stages of breast cancer, and most especially in the medical field, where those methods are widely used in diagnosis and analysis for decision-making. We focused on boosting classification models and evaluated the performance of XGBoost, AdaBoost, and Gradient Boosting. Our goal is to achieve higher accuracy by using boosting classifiers with hyperparameter tuning for the prediction of breast cancer stages, precisely the distinction between "Benign" and "Malignant" types of breast cancer. The Wisconsin breast cancer dataset is employed from the UCI machine learning database. The performance of our model was evaluated using metrics such as accuracy, sensitivity, precision, specificity, AUC, and ROC curves for various strategies. After implementing the model, this study achieved the best model accuracy, and 98.60% was achieved on AdaBoost.