Malicious Code Detection Technology Based on A3C Algorithm

摘要

In the face of the explosive growth of malicious code, this paper proposes a malicious code detection technology based on the A3C algorithm. The technology constructs a malicious code anti-detection adversarial model through the A3C algorithm of the reinforcement learning. It can automatically generates anti-detection malicious samples that bypass the intelligent model on the existing malicious sample set, constructs the anti-detection sample training data set, uses the anti-detection sample training data set to iteratively train new intelligent detection models, continuously builds a reinforcement learning intelligent detection model with continuously enhanced capabilities.