期刊名称：International Journal of Security and Its Applications
印刷版ISSN：1738-9976
出版年度：2016
卷号：10
期号：6
页码：75-86
DOI：10.14257/ijsia.2016.10.6.09
出版社：SERSC
摘要：Malicious code detection is one of the important missions of malicious code analysis. Current researches on the detection of malicious code mostly focused on single classifier, whereas the single classifier is not suitable for the detection based on features of different types. We utilized multi-classifiers ensemble based on fuzzy integral to improve the accuracy of the detection framework. A framework based on the Choquet fuzzy integral was proposed to fuse the analysis results of the base classifiers with different features. And the genetic algorithm was used to obtain the fuzzy measure. Finally, the result of Choquet fuzzy integral was compared to a threshold predefined to determine the maliciousness of binary code. Experiment showed that the framework proposed in this paper could be used to determine the maliciousness of binary code more accurately.
