摘要
为了提高以Android平台为主的互联智能产品恶意检测效率,降低恶意攻击风险,利用朴素贝叶斯算法的简单快速特点,引入基因表达式编程(Gene Expression Programming,GEP)算法优化提取出的Android APK静态信息样本数据中缺失属性的填充,一定程度上消除了样本数据属性值缺失对分类检测算法的影响。利用优化后的样本数据,对Android平台中的恶意应用进行静态检测,实验结果表明,该算法在Android恶意应用静态检测方面上,获得较好分类结果,能够高效检测Android恶意应用,降低智能产品的安全威胁。
To improve the efficiency of malicious detection and reduce the risk of malicious attacks on smart terminals based on Android,using the simple and fast characteristics of the Naive Bayes algorithm,the gene expression programming algorithm(GEP)is introduced to optimize the filling of the missing attributes in the extracted Android APK static information sample data,which eliminates the influence of the missing attribute values of the sample data on the classification detection algorithm.Using the optimized sample data,the malicious applications in Android platform are statically detected.The experimental results show that the algorithm obtains better classification results in the static detection of Android malicious applications,and can efficiently detect Android malicious applications and reduce security threats of intelligent products.
作者
罗锦光
杨鸣坤
苏锦
LUO Jinguang;YANG Minkun;SU Jin(School of Artificial Intelligence and Information Engineering,Guangxi Electrical Polytechnic Institute,Nanning Guangxi 530007,China;College of Computer Science and Engineering,Guilin University of Aerospace Technology,Guilin Guangxi 541004,China)
出处
《信息与电脑》
2021年第16期62-66,共5页
Information & Computer
基金
广西高校中青年教师科研基础能力提升项目“基于GEP-CPN的可信网络终端行为关键技术研究”(项目编号:2020KY41018)。
关键词
静态检测
恶意应用
机器学习
分类检测
static detection
malicious applications
network terminal behaviors
the classification detection
