Detection of malware based on their behavioral analysis using aggregate methods

Malware is emerging every day in a new form with new capabilities. But in the meantime, covert malware is trying to keep itself out of the sight of intrusion detection systems. This type of malware can continue to operate for years without being detected, stealing information from individuals, companies and even countries, causing irreparable damage. Therefore, timely detection of this type of malware is even more important. The aim of this study is to investigate the performance of the proposed method on standard malware datasets. First, the results of each processing step on the data are reviewed, and finally, the results of the proposed algorithm will be tested on the data and compared with other works. In this research, the effective features in detecting malware are determined using their behavioral analysis. The accuracy of malware detection has also been increased by using the cumulative random forest classifier. The evaluation criteria of the proposed algorithm are the accuracy, precision, sensitivity and F-criteria in classifying the classes in the data. The evaluation criteria in the proposed algorithms are compared with other methods and the results of these comparisons are presented in tables. The results show that malware detection using the proposed method has high accuracy, precision, sensitivity and F-criteria compared to other methods.