Evaluation of ML-based Breast cancer diagnosis using AWS SageMaker

This study explores the application of various machine learning methods for diagnosing breast cancer through binary classification. Algorithms such as Support Vector Machines (SVM), Naive Bayes, k-Nearest Neighbors (KNN), and Classification and Regression Trees (CART) were employed to classify tumors as benign or malignant based on data from the Breast Cancer Wisconsin Dataset. The objective is to identify the most effective model for accurate breast cancer detection by evaluating these algorithms using metrics such as accuracy and precision. Results indicate that CART, KNN, and Naive Bayes achieved approximately 92% accuracy, whereas SVM outperformed with an accuracy exceeding 99.1%. This study demonstrates the potential of SVM and other machine learning models to enhance early breast cancer detection, which could significantly improve patient survival rates. The findings highlight the critical role of machine learning in advancing medical diagnostics and improving healthcare outcomes.