KLASIFIKASI MALWARE ANDROID DENGAN MENGGUNAKAN METODE CATBOOST ALGORITMA

In 2008, Android was introduced as a popular open source project due to its customizability and low hardware requirements. Mid-2021 statistics from GlobalStat Counter shows that Android dominates the mobile operating system market with 72.74%. Despite its popularity, Android is becoming a target for malware attacks in the context of cyber crime. This problem prompted this research to be carried out with the aim of identifying and classifying Android malware which is continuously developing by applying machine learning logic, especially using the methodCatBoost. This method was chosen based on its effectiveness in previous research which has been proven to provide high accuracy. Performance evaluation involves comparisons betweenCatBoost and several previous researchers' methods, inclKNN (K-Nearest Neighbors), SVM (Support Vector Machine), LR (Logistic Regression), RF (Random Forest), ET (Extra Trees), XG (XGBoost), AB (Adaboost), and BG (Bagging), using common metrics such asValidation Accuracy, Detection Accuracy, and F1-Score. The research results show thatCatBoost managed to achieveValidation Accuracy amounting to 96.66%,Detection Accuracy 96,87%, andF1-Score of 96.81% puts it in a competitive position with most other methods, exceptRF (Random Forest). CatBoost's consistent superiority in this comparison shows its potential as an effective and consistent solution in Android malware detection and classification.