Breast Cancer Diagnosis using Soft Voting Classifier Approach
Archana Singh, Kuldeep Singh Kaswan · 2024
Breast cancer (BC), the leading type of cancer in women, is very important since the necessity of choosing exactly between benign and malignant cases requires highly accurate and precise diagnostic tools. Therefore, it is crucial not only to diagnose it at an early stage but also to use that result for the intended treatments. One of the aims of this study is to forecast breast cancer in an early and accurate manner. The proposed method in our research is the use of a soft voting classifier for automatic assessment of malignancy or benignancy of breast cancer using three ML algorithms: logistic regression, SVM, and decision tree. The Breast Cancer Wisconsin dataset (Original) with 699 instances is used to test and evaluate the proposed approach. The data is balanced using the random oversampling method to minimize the bias. The methodology that is proposed, gives 0.9708 accuracy, 0.9821 precision, 0.9483 recall, and an F-1 score of 0.9649 with an AUC of 0.9678.