Bio-Inspired Computational Paradigm for Feature Investigation and Malware Detection: Interactive Analytics

摘要: Recently, people rely on mobile devices to conduct their daily fundamental activities. Simultaneously, most of the prefer with Android operating system. As demand expands, deceitful authors develop malware compromise for private and money purposes. Consequently, security analysts have static dynamic analyses counter violation. In this paper, we adopt analysis which only requests minimal resource consumption rapid processing. However, finding a minimum set features in are vital because it removes irrelevant data, reduces runtime machine learning detection dimensionality datasets. Therefore, investigate three categories features, permissions, directory path, telephony. This investigation considers frequency as well repeatedly used each application. Subsequently, study evaluates proposed bio-inspired classifiers artificial neural network (ANN) category signify usefulness ANN type uncovering unknown malware. The multilayer perceptron (MLP), voted (VP) radial basis function (RBFN). Among all these classifiers, outstanding outcomes acquire is MLP, achieves 90% accuracy 87% true positive rate (TPR), 97% our Bio Analyzer prediction