Efficient Breast Cancer Dataset Analysis Based on Adaptive Classifiers

Abstract

Many algorithms have been used to diagnose diseases, with some demonstrating good performance while others have not met expectations. Making correct decisions with the minimal possible errors is of the highest priority when diagnosing diseases. Breast cancer, being a prevalent and widespread disease, emphasizes the importance of early detection. Accurate decision-making regarding breast cancer is crucial for early treatment and achieving favorable outcomes. The percentage split evaluation approach was employed, comparing performance metrics such as precision, recall, and f1-score. Kernel Naïve Bayes achieved 100% precision in the percentage split method for breast cancer, while the Coarse Gaussian support vector machines achieved 97.2% precision in classifying breast cancer in 4-fold cross-validation.