OPTIMAL MACHINE LEARNING CLASSIFIER IN MOBILE MALWARE DETECTION

As the development of mobile device technology continues to grow, more mobile apps have been launched to make everyday tasks even easier. As a forum for users to do social networking, online banking, online shopping, web surfing, and many other things, Smartphone apps have also been used. Most of these apps enable users to have private credentials, which can theoretically be sniffed and abused by cybercriminals. In minimizing mobile malware, various detection methods have been implemented, but the malware author has its own method of overcoming the method of detection. Therefore, to protect Smartphone users against malicious threats, an improved mobile malware detection technique is required. On this basis, this paper discusses the study of mobile malware detection and proposes the optimal classifier approach for machine learning. Based on previous studies, many machine learning classifier techniques are selected and their performances are analyzed. Based on the contribution made to enhancing the True Positive Rate (TPR), False Positive Rate (FPR) in the classification of benign and malicious mobile malware, the machine learning classifier methods are evaluated.