Breast Cancer Detection Using Machine Learning Algorithms

Abstract

Breast cancer is the most common type of cancer among women worldwide and has the highest mortality rate among women. As early diagnosis is important in cancer, early diagnosis in breast cancer significantly reduces the death rate. Thus, early detection of breast cancer significantly increases the chances of survival. Early diagnosis of breast cancer can significantly increase the chances of survival, as it can encourage timely clinical treatment. In this study, the data quality of the Breast Cancer Wisconsin (Diagnostic) dataset, which includes metric data extracted from the biopsy piece with various data mining methods was increased and the patient's breast cancer was classified as benign or malignant with machine learning algorithms. When we compare the developed machine learning algorithms; K-Nearest Neighbor algorithm showed higher performance than other machine learning algorithms with 99.3% accuracy, 98.9% precision, 100% recall and 99.4% f1-score values. The second most successful model on the test set is Support Vector Machine and Logistic Regression.