Detecting Metamorphic Virus Using Hidden Markov Model and Genetic Algorithm

摘要: Metamorphic viruses dodges the classical signature-based detection system by modifying internal structure without compromising on original functionality. To solve this problem, some machine learning technique, like Hidden Markov model (HMM) and Neural Network are can be used. HMM is a state where each observes input data with appropriate observation probability. learns statistical properties of antivirus features rather than signatures relies such statistics to detect same family virus. Each when trained variants that generated metamorphic engine so similar high But, in order make viruses, there three basic criteria needs satisfied. Generally most based techniques, Baum-Welch method used for solving one problems, i.e, estimating parameters corresponding given an output sequence. In paper, we have Genetic Algorithm problem. The selection algorithm over conventional Baum- Welch lies non-linearity genetic algorithm. algorithm, being linear nature, suffers from local optima which tried overcome using our scheme.