Penerapan K-Means untuk Pengelompokan Data Heart Failure Clinical Records

Abstract

This study aims to classify brain tumors using the Decision Tree algorithm based on numerical features extracted from MRI images. The dataset includes first-order statistical features and second-order texture features derived using the Gray Level Co-occurrence Matrix (GLCM) method. The research process consists of dataset collection, preprocessing, data splitting, model construction, and performance evaluation, all performed using RapidMiner Studio. Data were divided into 90% training and 10% testing to optimize pattern learning and ensure representative evaluation. The Decision Tree model was constructed using default parameters with pruning to prevent overfitting. The evaluation results show that the model achieved an accuracy of 97.46%, supported by high precision and recall values for both tumor and non-tumor classes. The confusion matrix indicates that the model correctly classified the majority of instances, demonstrating reliable predictive performance. The resulting decision tree reveals that Entropy, Homogeneity, Energy, Skewness, and Mean are influential attributes in distinguishing tumor from non-tumor cases. Overall, the findings indicate that the Decision Tree algorithm provides a highly interpretable and effective approach for brain tumor classification and can be utilized as a decision-support tool in medical diagnostics