ITM Web of Conferences (Jan 2024)
Enhanced machine learning models for predicting breast cancer: Healthcare system
Abstract
Currently, breast cancer is a popular illness that can lead to many consequences, with the most severe outcome being death rates. Therefore, there is a pressing requirement for a diagnostic tool that can aid healthcare professionals in early detection of the illness and provide required lifestyle modifications to prevent its development the possibility of developing cancer at a young age has also been significantly enhanced by environmental alterations in our daily existence. This analysis aimed to accurately classify features into either malignant or benign classes. The suggested methodologies and classifying systems were applied to the Wisconsin Diagnostic Breast Cancer (WDBC) and Breast Cancer Coimbra Dataset (BCCD) datasets. Conventional performance measures, such as (KNN, SVM, ensemble classifier (EC), and logistic regression (LR)) methods, were utilized to evaluate the efficacy and time of training for each classifier. The diagnostic power of the models was enhanced by our DET (Diagnostic Enhancement Technique). Specifically, the polynomial SVM achieved an accuracy of 98.3%, LR (Logistic Regression) reached 97.04%, KNN (KNearest Neighbors) achieved 96.3%, and EC (Ensemble Classifier) achieved 96.6% accuracy with the dataset is called WDBC. In addition, in this study, there’s just make a comparative analysis of the findings in relation to the accuracy of the outcomes of prior research. The implementation process and results can assist clinicians in adopting an efficient prototype for functional comprehension and forecast of breast cancer (BC) tumours.