Malicious code homologous judgment method based on deep learning

Abstract

The invention provides a malicious code homologous judgment method based on deep learning. The method comprises the following steps: disassembling the to-be-judged malicious code by using an IDA tool,obtaining the core binary content therein through regular expression matching, and removing useless information at the same time; receiving the binary content as the input, mapping the binary contentas the malicious code image by using the malicious code visualization algorithm; training a deep learning model convolutional neural network by adopting a sample set composed of the malicious code image and a label value, thereby obtaining a mature judgment model; and receiving the to-be-judged malicious code as the input to accomplish the homologous judgment. The homogeneous judgment task is converted into an image classification task through the malicious code visualization algorithm, and the available malicious code homogeneous judgment method is realized by combining the deep learning judgment model. And the homogeneous judgment technology with higher judgment accuracy rate in comparison with the existing system is realized.

Description

technical field [0001] The present invention relates to the technical field of Internet information security. Specifically, it relates to a method for judging the same origin of malicious codes based on deep learning. Through a malicious code visualization algorithm, the same-origin judging task is converted into an image classification task, combined with a deep learning judging model , which implements a usable malicious code homology determination method. Background technique [0002] With the advent of the Internet age and the rapid development of information technology, while providing people with various conveniences, it also brings potential information security issues, especially network security issues, which threaten the security of user information and property, and the proliferation of malicious codes is one of them. Malicious code refers to the implementation code of all software carrying malicious attacks. It is a program code that violates the security policy...

AI Technical Summary

Problems solved by technology

[0005] At present, there is no description or report of the similar technology of the present invention, and no similar data at home and abroad have been collected yet.

Images