Model Selection for Predicting Breast Cancer using Supervised Machine Learning Algorithms

Abstract

Breast Cancer is the most common malignancy in women affecting 2.1 million women every year and causing the maximum number of deaths in women due to cancer. It occurs as a result of the unusual development of cells in the breast tissue, which is generally referred to as a Tumor. A tumor does not signify cancer. It may be not cancerous (benign), pre-cancerous (pre-malignant), or cancerous (malignant). Various types of tests such as mammograms, MRIs, ultrasound, and biopsy are frequently used to identify breast cancer. Early detection and treatment will help to improve breast cancer outcomes as well as survival. Therefore, this paper consists of a relative study of the breast cancer prediction using different supervised machine learning algorithms like Logistics Regression, K-Nearest Neighbors, Decision Tree Classifier, Gaussian NB, and Support Vector Machine on the UCI repository dataset. Concerning the performance of all the models, the accuracy score, precision, recall, and F-score of each model have been compared. After using various models, we got to see that Logistic Regression is a well-suited algorithm for Breast cancer prediction and came up with better accuracy and other performance indices as compared with other models.