Comparison-Based Study to Predict Breast Cancer: A Survey

Abstract

Cancer is utmost dangerous disease that leads to death stage if not cured on time. Breast cancer is the second most common disease after lung cancer in women. Therefore, its early detection is of utmost importance. Machine learning plays an important role to predict breast cancer in the early stages. In this paper, the authors present a comparison study to predict breast cancer on the Breast Cancer Wisconsin Diagnostic dataset by applying six different machine learning algorithms such as CART, logistic regression, support vector classifier, hard voting classifier, Extreme Gradient Boosting, and artificial neural network. Authors have used various metrics for model evaluation keeping accuracy as one of the most important factors since higher accuracy models can help doctors to better detect the presence of breast cancer.