A preliminary study for Malware detection across various platforms Available to Purchase

Malware is one of the major threats that can attack the security of many platforms, including Android and Windows. Due to the widespread use of these two platforms, applications and malware targeting both operating systems have increased dramatically. Malware is coded so cleverly that identifying it has become extremely complex, yet traditional signature-based methods for detecting malware have become practically useless. In order to find a suitable solution against malware threats, a machine learning-based malware detection system was proposed on Android and Windows platforms using two data sets, the first for Android and the second for Windows. The datasets are pre-processed, and then feature selection using the feature ranking algorithm is applied to produce different feature-specific datasets. The selected features are then fed into four machine learning classifiers to be trained (RF, DT, SVM, and KNN). The trained models are then evaluated using several performance metrics (precision, precision, recall, and F1 score). The results indicate that the RF algorithm achieved the highest results on both systems, as it was 98.5% on the Android data set and 99.8% on the Windows data set.