The Use of the Binary Bat Algorithm in Improving the Accuracy of Breast Cancer Diagnosis

Introduction: The early diagnosis of breast cancer as prevalent cancer among women, is a necessity in the research on cancers since it could simplify the clinical management of other patients. The importance of the classification of breast cancer patients into high- or low-risk groups has led research groups in the biomedical and informatics departments to evaluate and use computer techniques such as data mining. To date, various methods have been used for breast cancer diagnosis which has shown unfavorable accuracy due to issues such as computational complexities and prolonged implementation. Methods: The present study aimed to apply the feature selection method based on the binary bat algorithm (BBA) to increase the accuracy of the breast cancer diagnosis. Feature selection is carried out to select the most important features from a dataset. We applied the naïve bayes (NB), support vector machine (SVM), and J48 algorithms in MATLAB software; based on the dataset obtained from Wisconsin to evaluate the accuracy, sensitivity, and diagnostic criteria of the proposed model. Results: The BBA had 99.28%, 96.43%, and 92.86% accuracy in SVM, NB and J48 algorithms, respectively. Conclusions: According to the results, the feature selection technique, along with the BBA and SVM, yielded the most accurate results regarding breast cancer detection.