A Comparative Analysis of Breast Cancer Classification using Ensemble and Multi Classifier Algorithms

To help and monitor the patients, the initial diagnosis and type prediction of cancer should become mandatory in cancer research. Machine learning (ML) approaches are being studied and evaluated by several research teams from the biomedical sector due to the relevance of classifying cancer patients into two risk groups. It has been suggested that base classifiers be used in conjunction with ensemble methods to predict breast cancer. This work, provide detailed forecasts using data on breast cancer in a novel setting. In order to create precise predictions, this research investigates the many classification-based data mining techniques. Additionally, this study evaluates the dataset using a variety of classifiers to forecast which model will perform the best. The cancer dataset worked upon is the Wisconsin Dataset with 569 instances. This Dataset is given to classifiers like SVM, J48, REP tree and Random Forest after pre-processing and cross validation is applied later, new models are tested subsequently. The results were assessed across multiple parameters, and the analysis indicates that among all the classifiers, the REP tree focusing on node caps delivers the most in-depth predictions and achieves the highest accuracy. This model outperforms other approaches, including Boosting with Random Forest (node caps), Boosting with Random Forest (INV nodes), and Boosting with J48, in terms of yielding superior and precise results.