IMCDCF: An Incremental Malware Detection Approach Using Hidden Markov Models

Abstract

Dynamic malware analysis has become popular because it allows analysts to observe the behavior of running samples, facilitating improved decisions for malware detection and classification. With the increasing number of new malware, there is a growing need for an automated malware analysis engine that can accurately detect malware samples. In this paper, we briefly introduce the malware detection and classification approaches. Furthermore, we introduce a new malware detection and classification framework that works specifically in the dynamic analysis setting, namely Incremental Malware Detection and Classification Framework, or IMDCF. In this paper, we present a novel framework designed specifically for the dynamic analysis setting, named the Incremental Malware Detection and Classification Framework (IMDCF). IMDCF provides a end-to-end solution for general-purpose malware detection and classification with 96.49% accuracy and simple architecture.