Machine Learning Approaches for Investigating Breast Cancer

Abstract

ABSTRACT: This study aims to predict whether the case is malignant or benign and concentrate on the anticipated diagnosis; if the case is malignant, it is advised to admit the patient to the hospital for treatment. The primary goal of this work is to put together models in two distinct datasets to predict breast cancer more accurately, faster, and with fewer errors than before. Then contrast the techniques that produced datasets with the highest accuracy. In this study, the datasets were processed using Support Vector Machine, Logistic Regression, Decision Tree, K-Nearest Neighbours, Artificial Neural Network, Nave Bayes, Stochastic Gradient Descent (SGD),Gradient boosting classifiers(GBC), Stochastic Gradient Boosting (SGB), Extreme Gradient Boosting (XGBoost),and Random Forest. Two datasets—the Wisconsin Diagnostic Breast Cancer dataset and the Breast Cancer dataset—are used to test these methods. to evaluate the findings and choose the algorithm that is more adept in predicting breast cancer. Seven algorithms that operate on both datasets in the AI platform were used to build the article. Breast cancer prediction has gotten much harder because so many people die from the disease in its early stages. Consequently, using two real-time datasets, one for Wisconsin diagnosis and the other for research on breast cancer. The same methods are applied to both datasets, and it is found that SVM provides the best accuracy in the shortest time and with the lowest error rate.