Breast Cancer Detection: An Evaluation of Machine Learning, Ensemble Learning, and Deep Learning Algorithms

Breast cancer is a prevalent and potentially life-threatening disease that affects a significant number of individuals globally. Early detection and precise diagnosis can help to control the death rate. With the help of Wisconsin Breast Cancer Diagnostic (WBCD) dataset, this paper gives a detailed comprehensive comparison analysis of machine learning, ensemble learning, and deep learning for breast cancer diagnosis. The study’s focus is on determining the efficacy of various algorithms in identifying breast cancers as benign or malignant. The study’s goal is to enhance the cross-validation accuracy and precision of breast cancer diagnosis by examining the effectiveness of these algorithms. This study investigates a wide range of classification algorithms, such as Logistic regression, Gaussian Naïve bayes, Support vector machine, Random forest, XGBoost, Adaboost and Deep neural network. This study analyzes the 5 fold cross-validation accuracy, precision, recall, f1-score, roc-auc score and roc curve of different models using considerable experimentation. This comprehensive comparative analysis also revealed that the DL_ANN model performed the best of all, with a validation accuracy of 98.50%. The findings presented in this paper hold promise for reducing the incidence of breast cancer by developing a predictive system based on machine learning the disease can be detected early.