Static Heuristics Classifiers as Pre-Filter for Malware Target Recognition (MATR)
Anuj Lohani · American Journal of Networks and Communications · 2015
Now a day’s malware are one of the major threats to computer information system. The current malware detection technologies have certain significant limitations on their part. Different organizations which deal with the protection of sensitive information may face the problem in identifying recent malware threats among millions and billions of benign executables using just signature-based antivirus systems. Currently for frontline defense against malware, signature-based antivirus products are used by organization.In the undergoing project, we proposed a detection approach by using static heuristics in MATR for malware in PE (portable executable) files. The project suggestslarger performance-based malware target recognition architecture that at present use only static heuristic features.Results of the experiments show that this architecture achieves an overall test accuracy of greater than 98% againstmalware set collected from various operational environments, while most antivirus provide detection accuracy of only 60% at their most sensitive configuration [1]. Implementations of this architecture enables benign executables to be classified successfully to some extent providing enhanced awareness of operators in hostile environments it also enable detection of unknown malware. We are to show the performance of Bagging and AdaBoostensemble.