基于改进决策树分类的Android恶意软件检测

Android Malware Detection Based on Improved Decision Tree Classification

针对目前Android手机恶意软件越来越多的问题,本文在现有研究的基础上,设计了一个Android恶意软件检测框架。该框架通过提取Android应用程序的特征属性,结合Fisher Score、信息增益和卡方检验三种特征选择方法,对属性特征进行预处理,然后利用恶意检测模块中的改进决策树算法对软件进行分类。通过实验仿真,结果表明使用该检测框架检测恶意软件具有较低的误报率和较高的精确度。

At present, the mobile phone for Android is facing more and more malicious threats, this paper designs an Android malware detection framework. A malicious detection framework is implemented which adopts Android application extraction feature, Fisher Score and information gain and chi-square test as the feature selection method. The attributes are pre processed, and then the improved decision tree algorithm is used to classify the software. The simulation results show that the malware detection framework own lower false alarm rate and higher accuracy.