ABSTRACT
Malware, or any malicious software, has the ability to attack a computer, damage infrastructure, disrupt computer networks, and take data from users without their consent. Ransomware, worms, trojan viruses, spyware, and adware are examples of common malware. Our platform enables the utilisation of several machine learning techniques to effectively compare malware and clean files, with the goal of reducing false positives and maximising detective efficacy. An edge computing system is suggested in order to address the issues of efficiency, low latency, high processing volume, and privacy concerns. In this paper, we proposed a flexible framework that works with data splitting as a Testing and Training Dataset after working on data pretreatment initially. By classifying the data using SVM, Random Forest, and decision tree algorithms, along with the software quality prediction dataset and other sources, this system will improve the accuracy of the classification findings. The ideas behind this framework underwent a scaling-up process after successful testing on medium-sized datasets of malware and clean files, enabling us to handle incredibly large datasets of malware and clean files.
