- 期刊
Android Applications Malware Analysis Using Machine Learning Based Approach
摘要
In the rapidly evolving domain of mobile security, effective detection of Android application malware is a critical challenge. In the emerging security industry, one of the biggest concerns is detecting malware in Android applications. This study explores how machine learning techniques can be used to identify malware in Android apps using a dataset called Drebin malware. We conducted an evaluation of machine learning models, including Random Forest, Naive Bayes, Artificial Neural Network, Perceptron, Sequential Neural Networks (NN), K-Nearest Neighbors (K-NN) Logistic Regression, and Support Vector Machines (SVM) with various kernels, like Radial Basis Function (RBF), Polynomial (Poly), and Linear as well as Decision Trees. The performance of these models was assessed based on accuracy, precision, recall, and F1-score. Our findings revealed that Random Forest and Artificial Neural Network models significantly outperformed the others, achieving accuracy rates of 98.77% and 98.57%, respectively. These models also outperformed others in terms of precision, recall and F1-score. While the Naive Bayes model showed efficiency compared to others, SVM with RBF kernel and Logistic Regression also demonstrated performance. This research highlights the capabilities of advanced machine learning algorithms such as Random Forest and ANNs when it comes to detecting Android malware within the Drebin dataset. The findings provide insights for enhancing cybersecurity measures, against the challenges presented by Android malware. These insights provide a valuable contribution to the field of cybersecurity, underscoring the effectiveness of machine learning in combating the sophisticated and evolving threats in Android malware.