An Ensemble Based Machine Learning Model for Android Malware Detection

• Author(s): Baffa Sani Mahmoud; Prof. Rashid Husain; Assoc. Prof. Muhammad Hassan
• Paper ID: 1711462
• Page: 1207-1217
• Published Date: 27-10-2025
• Published In: Iconic Research And Engineering Journals
• Publisher: IRE Journals
• e-ISSN: 2456-8880
• Volume/Issue: Volume 9 Issue 4 October-2025

Abstract

The rapid growth of the Android ecosystem has been accompanied by an alarming increase in sophisticated malware, including banking Trojans, spyware, and ransomware. Traditional signature-based detection techniques are insufficient against obfuscation and zero-day attacks, highlighting the urgent need for adaptive detection mechanisms. This study aims to develop and evaluate an ensemble-based machine-learning model to enhance the detection of Android malware using the Andmaldataset. Recursive Feature Elimination (RFE) with a Decision Tree Classifier was employed to select the 20 most relevant features from the dataset. Five supervised classifiers Random Forest, Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Logistic Regression, and Decision Tree were trained and evaluated. Additionally, three ensemble-learning techniques (Bagging, Boosting, and Stacking) were implemented to improve robustness and reduce false negatives. Among individual classifiers, SVM achieved the highest accuracy of 96.51%, while Random Forest recorded the strongest AUC score (0.9918). Ensemble methods outperformed individual classifiers, with Boosting yielding the highest accuracy (98.51%) and recall (96.32%), and Bagging achieving the best AUC (0.9930). Stacking also demonstrated stable and competitive performance across all metrics. The results confirm that ensemble learning significantly improves Android malware detection over single classifiers. Boosting and Bagging emerged as particularly effective strategies, offering strong accuracy and robustness against evolving malware threats.

Keywords

Android Malware, Machine Learning, Ensemble Learning, Bagging, Boosting, Stacking, Malware Detection

Citations

IRE Journals:
Baffa Sani Mahmoud, Prof. Rashid Husain, Assoc. Prof. Muhammad Hassan "An Ensemble Based Machine Learning Model for Android Malware Detection" Iconic Research And Engineering Journals Volume 9 Issue 4 2025 Page 1207-1217

IEEE:
Baffa Sani Mahmoud, Prof. Rashid Husain, Assoc. Prof. Muhammad Hassan "An Ensemble Based Machine Learning Model for Android Malware Detection" Iconic Research And Engineering Journals, 9(4)