PENGGUNAAN ALGORITMA RANDOM FOREST UNTUK DETEKSI MALWARE BERDASARKAN POLA PERMISSION PADA APLIKASI ANDROID KOMUNIKASI

Android is the most widely used mobile operating system, holding a market share of 72.04% in the first quarter of 2024. Its popularity makes it vulnerable to malware threats, including the misuse of permissions by applications. According to a Kaspersky survey, the number of malware attacks on mobile devices increased significantly in 2023, reaching 33,790,599 attacks. This study aims to analyze malware detection based on permission patterns in Android communication category applications using the Random Forest algorithm. The dataset used consists of 3,437 records obtained from Mendeley Data. The method applied is CRISP-DM, which includes the stages of Business Understanding, Data Understanding, Data Preparation, Modeling, and Deployment. The research was conducted under three different data split ratio scenarios to determine the best results. Evaluation results showed that the best scenario was a 70% training and 30% testing data split, using model parameters of n_estimators=100, max_depth=None, and random_state=45, which yielded an accuracy of 89%. The model was then implemented on a website to analyze APK files and classify malware risk levels into three categories: low, medium, and high.