Breast Tumor Detection using Machine Learning Boosting Classifiers

Abstract

Breast cancer is most commonly found in women and the second greatest cause of death worldwide. As breast cancer is detected early, the ratio of survival rate increases because better therapy may be provided. ML algorithms play a critical role in the early detection of breast cancer. In this study, we purposed a Novel method that increases the accuracy and performance using these three different classifiers: Gradient Boost (GB), Ada Boost (ABC), and Extreme Gradient Boost (XGB). On the Public dataset WBC, we evaluated and compared the classifiers’ performance and accuracy. Because the chance of examples belonging to the majority of the class is relatively high, algorithms are far more likely to categorize new observations into the majority class in the classification phase. We address such a situation that True positive, false positive, precision, recall, F1 score, and accuracy are all used to evaluate the efficiency of each classifier. Experiments demonstrate that utilizing a boosting classifier improves the performance, with the Gradient Booster (GB) outperforming others in the WBC dataset.