Review on Design Android Malware Family Classification Based On Hybrid Forensic Analysis Tool Using Deep Learning

Android is widely used by both users and malicious actors. Attackers on Android are drawn to the large number of Android users. Our detection techniques should require an update as well, given the ongoing expansion of Android malware's variety and assault techniques. Fewer studies concentrate on dynamic features than the majority, which are based on static features. We are addressing the gap in the literature in this research by employing System calls and A malware substring is initially made up of a series of several substrings, each of which is termed a pixel and is 8 bits long. In the following phase, the 8-bit substring is transformed into a decimal number between 0 and 255. Moreover, every virus substring was changed. transformed into a vector in one dimension and then into a two-dimensional matrix with a defined breadth. It was dubbed a "malicious code matrix."The two-dimensional grayscale image is this matrix. In this research, we proposed a clustering algorithm-based approach for classifying Android malware. Testing the suggested rule-based clustering technique on a dataset with the best accuracy and lowest mean absolute error by 98.12%, respectively, yields better results.