Abstract:The applications which static analysis methods can not accurately identify and determine,such as the high degree of source code confusion or those wihch use dynamic code loading technology,we extracting the network traffic generated during the Internet connection.Then the information gain algorithm is used to select the discrimination features and set up improved nave Bayes classifier with natural logarithm of multiplier and Laplace calibration.After ten-fold cross validation method,the improved Bayes model can reduce the time complexity and achieve 93% accuracy.Comparing with the method based on privilege features,the improved nave Bayes classifier based on traffic features has better classification performance,at the same time,this method also provides a new way to detect android malwares,accurately.