A Malware Identification and Detection Method Using Mixture Correntropy-Based Deep Neural Network

Abstract

With the rapid development of CPS technology, the identification and detection of malware has become a matter of concern in the industrial application of CPS. Currently, advanced machine learning methods such as deep learning are popular in the research of malware identification and detection, and some progress has been made so far. However, there are also some problems. For example, considering the existing noise or outliers in the datasets of malware, some methods are not robust enough. Therefore, the accuracy of classification of malware still needs to be improved. Aiming at it, we propose a novel method thought the combination of correntropy and deep neural network (DNN). In our proposed method for malware identification and detection, given the success of mixture correntropy as an effective similarity measure in addressing complex dataset with noise, it is therefore incorporated into a popular DNN, i.e., convolutional neural network (CNN), to reconstruct its loss function, with the purpose of further detecting the features of outliers. We present the detailed design process of our proposed method. Furthermore, the proposed method is tested both on a popular benchmark dataset and a real-world malware classification dataset, to verify its learning performance.