A New Method of Virus Detection Based on Maximum Entropy Model

摘要

The paper presents a new method for detecting virus based on the Maximum Entropy Model. This method is also used to detect unknown viruses. Maximum Entropy is a machine learning method based on probability distribution and has been successfully applied in the classification problem. From the background knowledge, we improve and apply this model to solve the problem of virus detection. In the training phase, virus samples in the virus warehouse are extracted their features and trained to create the Entropy Model. In the detection process, the Entropy Model is used to recognize virus based on the corresponding features of a checked file.