首页>
外国专利>
METHOD FOR FAMILY CLASSIFICATION BY WEIGHTED VOTING FOR ANDROID MALWARE LABELS RECORDING MEDIUM AND DEVICE FOR PERFORMING THE METHOD
METHOD FOR FAMILY CLASSIFICATION BY WEIGHTED VOTING FOR ANDROID MALWARE LABELS RECORDING MEDIUM AND DEVICE FOR PERFORMING THE METHOD
展开▼
机译:用于Android Malware标签的加权投票的家庭分类方法记录介质和用于执行该方法的设备
展开▼
页面导航
摘要
著录项
相似文献
摘要
An Android malicious app label weight voting-based family classification method includes: collecting characteristic information and label information from an Android malicious app; Learning a machine learning model using the collected characteristic information of the Android malicious app and the vaccine result of classifying the Android malicious app into families as a learning data set; extracting classification accuracy for each vaccine, which is a probability of accurately classifying a family for each vaccine through learning of the machine learning model; assigning a weight corresponding to the classification accuracy for each vaccine to the voting values of family candidates extracted using the label information collected from the Android malicious app; and ranking the family candidates by the weighted vote value to determine the family candidate having the maximum vote value as the final family. Accordingly, it is possible to classify a new family of Android malicious apps, and at the same time increase the accuracy of family classification based on label voting.
展开▼