Immune Principle and Neural Networks-Based Malware Detection

Detection of unknown malware is one of most important tasks in Computer Immune System (CIS) studies. By using nonself detection, diversity of anti-body (Ab) and artificial neural networks (ANN), this chapter proposes an NN-based malware detection algorithm. A number of experiments illustrate that this algorithm has high detection rate with a very low false positive rate. Aiming at automation detection malicious executables, the chapter proposes a novel malware detection algorithm (MDA) based on the immune principle and ANN. Extensive experiments show that the algorithm has a better detection performance than Schultz's method. The first goal is to verify the detection ability of the malware detection algorithm for malicious executables. The experimental results are all scaled with false positive rate (FPR) and detection rate (DR). The second goal is to calculate the probability of the reducing detection hole with the diversity of detectors.