| Abstrakt: |
Computer technology has become a necessity in human's life in various areas like online education, financial sector, entertainment, communication, etc. But computer security is vulnerable due to malware, which are the codes to damage the com- puter system. Some primary tools can detect the malware, known as malware detectors, whose quality depends on the techniques used in detectors. Malware analysis is the method of investigating the intention and practicality of the samples of malware like a worm, virus, trojan horse, etc. Static, dynamic, and hybrid approaches are used for malware analysis by various researchers. The machine learning techniques are most popular that employ these approaches. The machine learning approaches are also categorized as supervised, unsupervised, and reinforcement. Researchers employ one, two, or a blend of these approaches malware detection This research paper includes a study of these malware analysis techniques, and we analyze several machine learning algorithms and demonstrate the results obtained from the different machine learning algorithms. We compare outcomes of algorithms such as J48, Logistic Regression, and Random Forest. Moreover, we also employ a voting approach and show that Random Forest worksbetter than other algorithms. [ABSTRACT FROM AUTHOR] |
