Breast Cancer Detection and Classification: A Comparative Analysis Using Machine Learning Algorithms

Abstract

The risk of fatality from breast cancer is increasing exponentially as the population rises. It is the most typical kind of cancer and the major cause of death in women throughout the world. Early detection and treatment of breast cancer could significantly improve the prognosis and survival rate. It is an important phase in rehabilitation and medication since it can help patients receive prompt medical services. An automated disease detection technique that employs machine learning (ML) and deep learning (DL) techniques assist medical professionals in the diagnosis of diseases and provide a reliable, efficient, and faster response thereby minimizing the chance of death. This research aims to perform a comparison among ML and DL methods for breast cancer detection and diagnosis. The five most popular supervised ML techniques named support vector machine (SVM), decision tree (DT), logistic regression (LR), random forest (RF), K-nearest neighbor (KNN), and a DL technique were used for classification using cross-validation technique. The Breast Cancer Wisconsin (diagnostic) dataset has been used as a training set to evaluate and compare each algorithm's effectiveness and efficiency through classification accuracy, recall, specificity, precision, false-negative rate (FNR), false-positive rate (FPR), F1-score, and Matthews correlation coefficient (MCC). Experimental results show that random forest (tuned) outperformed all the other models with accuracy and F1-score of 96.66% and 0.963, respectively.KeywordsBreast cancer classificationMachine learningEarly diagnosisDisease predictionMedical data mining