A comparative assessment of malware classification using binary texture analysis and dynamic analysis

摘要: AI techniques play an important role in automated malware classification. Several machine-learning methods have been applied to classify or cluster into families, based on different features derived from dynamic review of the malware. While these approaches demonstrate promise, they are themselves subject a growing array counter measures that increase cost capturing binary features. Further, feature extraction requires time investment per does not scale well daily volume instances being reported by those who diligently collect Recently, new type extraction, used classification approach called binary-texture analysis, was introduced [16]. We compare this existing previously published. find that, while texture analysis is capable providing comparable accuracy contemporary techniques, it can deliver results 4000 times faster than techniques. Also surprisingly, texture-based seems resilient packing strategies, and robustly large corpus with both packed unpacked samples. present our experimental three independent corpora, comprised over 100 thousand These suggest could be useful efficient complement analysis.