Time and Computation Efficient Malicious Android Application Detection Using Machine Learning Techniques

Malware has become one of the major threats to information security in this rapid growth of internet applications. This has led the researchers to develop distinguished methods for detecting malware in this respect. To address the issue, machine learning techniques have proven itself efficient in detecting malware. But one of the major challenges is the reduction of attributes or components that are less important in malware detection process. Applying Principal Component Analysis (PCA) with other machine learning techniques, successful reduction of components is possible without any alternation in detection accuracy. In this paper, an approach based on PCA has been proposed which is time and computation efficient in detecting malware than the existing ADROIT approach that doesn't use PCA. Experimental results have also shown the best suited approach for further development in dynamic malware detection process.