Exploring Feature Extraction and ELM in Malware Detection for Android Devices

摘要: A huge increase in the number of mobile malware brings a serious threat to Internet security, as adoption rate device is soaring, especially Android device. variety researches have been developed defense malware, but users continuously suffer private information leak or economic losses from malware. Recently, large methods proposed based on static dynamic features analysis combining with machine learning methods, which are considered effective detect In this paper, we propose an framework feature extraction and neural network calssifier. framework, take use represent utilize extreme ELM algorithm learn network. We first extract then three different including principal component PCA, Karhunen-Loeve transform KLT independent ICA matrix into new spaces generate matrixes. For each matrix, construct En base classifiers by using ELM. Finally, Stacking method combine results. Experimental results suggest that detecting