Model Proposal for Malware Detection Using Deep Learning on Cell Phones with Android Operating System

Silvera D.; Molina P.; Ticona W. · Lecture Notes in Networks and Systems (2024) · DOI: 10.1007/978-3-031-70518-2_22

Resumen

In recent years, with the advancement of technology, the number of malware attacks has also increased, Peru being one of the countries with more cyberattacks of this type to electronic devices. Therefore, the main objective of the proposed study is to design a Malware detection model using Deep Learning techniques in cell phones with Android operating system. For this purpose, 5 phases have been carried out in the methodology to be able to recognize Malware anomalies and present error-free data comprising the following: Imported dataset, preprocessing, feature extraction, model implementation and evaluation. In addition, different machine learning models such as DT, RF, SVM and K-NN were developed and the results were evaluated using the metrics Accuracy, Precision, Recall and F1-score. For malware detection, RF presents the highest percentage in all the indicated parameters 89.23%, 87.59% and 88.90% and in Recall it is below K-NN with 85.84%. The RF model outperforms all the algorithms applied in the model, showing that better predictive results can be obtained. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.

Instituciones: Universidad ESAN; Universidad Tecnológica del Perú