[In-person]Intelligent Android Malware Detection in Real-World Environments Using Machine and Deep Learning Model

Intelligent Android Malware Detection in Real-World Environments Using Machine and Deep Learning Model

ID：149 Submission ID：119 View Protection：ATTENDEE Updated Time：2025-12-23 13:19:08 Hits：309 In-person

Start Time：2025-12-29 14:15 （Asia/Amman）

Duration：15min

Session：[S3] Track 3: Privacy, Security for Networks » [S3] Track 3: Privacy, Security for Networks

Presentation File

Tips: The file permissions under this presentation are only for participants. You have not logged in yet and cannot view it temporarily.

Abstract

This paper proposes a novel algorithm to detect the Android malware. The algorithm implements several models of machine and deep learning. The proposed method uses TUANDROMD to develop and assess these models; six models were implemented; namely, Gradient Boosting, Random Forest, Support Vector Machine, XGBoost, Multi-Layer Perceptron, and a custom deep learning model developed in PyTorch. The Random Forest classifier has the highest performance, with an AUC of 1.00 indicating that the algorithm is able to accurately classify samples with almost no errors, while the other models demonstrated strong predictive capabilities. In addition, the ROC curve indicates excellent performance in distinguishing between malicious and safe applications. These results give a good indication that machine learning and deep learning are robust in Android malware detection. This work contributes to the development of reliable, data-driven security solutions capable of addressing the evolving challenges in mobile threat detection.