<![CDATA[Abstrak: Klasifikasi tumor otak bertujuan untuk mengevaluasi dan membandingkan kinerja beberapa algoritma pembelajaran mesin dalam klasifikasi tumor otak menggunakan Citra MRI. Dalam penelitian ini, metodologi yang digunakan melibatkan pengujian algoritma tradisional seperti K-Nearest Neighbors (KNN), Naïve Bayes, Support Vector Machines (SVM), dan beberapa arsitektur Deep Learning seperti Neural Network Data yang digunakan dalam penelitian ini terdiri dari citra MRI otak yang telah dilabeli secara manual oleh ahli radiologi. Kami membandingkan kinerja algoritma berdasarkan beberapa metrik evaluasi, termasuk akurasi, presisi, recall, dan F1-score. Hasil penelitian menunjukkan bahwa algoritma berbasis Neural Network (0.99) secara signifikan mengungguli algoritma tradisional seperti KNN (0.98), Naive Bayes (0.97), dan SVM (0.98) dalam hal akurasi dan ketahanan terhadap variasi data. Namun, algoritma Neural Network dan metode ensemble menunjukkan kinerja yang kompetitif dengan keuntungan dalam hal interpretabilitas dan kecepatan pelatihan. Studi ini menyoroti keunggulan dan keterbatasan masing-masing algoritma dalam konteks klasifikasi tumor otak dan memberikan panduan praktis untuk memilih algoritma yang paling sesuai berdasarkan kebutuhan klinis dan karakteristik dataset. Penelitian lebih lanjut diperlukan untuk mengoptimalkan integrasi metode-metode ini dalam sistem pendukung keputusan klinis guna meningkatkan hasil diagnosis dan perawatan pasien===============================================Abstract:Brain tumor classification aims to evaluate and compare the performance of various machine learning algorithms in classifying brain tumors using MRI images. In this study, the methodology involves testing traditional algorithms such as K-Nearest Neighbors (KNN), Naïve Bayes, Support Vector Machines (SVM), and several deep learning architectures, including Neural Networks. The dataset used consists of brain MRI images manually labeled by radiology experts. We compared the performance of these algorithms based on several evaluation metrics, including accuracy, precision, recall, and F1-score. The results show that Neural Network-based algorithms (0.99) significantly outperform traditional algorithms such as KNN (0.98), Naïve Bayes (0.97), and SVM (0.98) in terms of accuracy and robustness to data variation. However, Neural Networks and ensemble methods demonstrated competitive performance with advantages in interpretability and training speed. This study highlights the strengths and limitations of each algorithm in the context of brain tumor classification and provides practical guidance for selecting the most suitable algorithm based on clinical needs and dataset characteristics. Further research is needed to optimize the integration of these methods into clinical decision support systems to enhance diagnosis and treatment outcomes for patients]]>
