Research on Android Malware Detection Technology Based on Improved SVM Algorithm

Abstract

In the training of the classifier, Android malware detection based on the traditional SVM algorithm often mixed with some feature vector, which increase the burden of the training device and reduce the computational power of the classifier, the actual effect of test and training time is often affected. In order to solve this problem, this paper proposed an Android malware detection technology based on improved SVM algorithm. After the feature vector extraction of normal software and simulated malware, the training sample set is obtained. When training the classifer, the algorithm optimizes and complements the training set to make the training data more reasonable. Experiments show that the Android malware detection system based on improved SVM algorithm is superior to the traditional SVM algorithm in the classification accuracy and detection rate of false positives, and with the increase of the training sample set greatly reduces the training time, this Android malware detection technology is more suitable for Android malware detection environment.