NALISIS PERBANDINGAN ALGORITMA MACHINE LEARNING KLASIFIKASI UNTUK DETEKSI TINGKAT KEGANASAN PENYAKIT KANKER PAYUDARA

This research focuses on implementing Machine Learning (ML), an Artificial Intelligence (AI) branch, to enhance breast cancer detection. The Synthetic Minority Over-sampling Technique (SMOTE) method is implemented for the first time in this study. The research compares various ML algorithms to improve the accuracy and efficiency of diagnosing the malignancy level of breast cancer, using the Wisconsin Breast Cancer dataset. Support Vector Machine (SVM) is identified as the best-performing algorithm, demonstrating high accuracy, a high Area Under the Curve (AUC), and good precision. Experimental results show the highest accuracy with an AUC value close to perfection (0.99). Furthermore, the identification of 10 factors causing breast cancer malignancy provides valuable insights. Despite contributing significantly to developing more effective detection methods, the research has two main limitations: reliance on a single dataset and the potential for expanding experiments by testing more classification algorithms. In conclusion, this study supports efforts for more effective breast cancer detection, hoping its findings can be applied more broadly.