Malware Classification using Machine Learning Approach

Abstract

Malware classification is essential for tracing the source of computer security threats. On the Internet, malware evolves at a rapid rate, and the bulk of undiscovered malware is developed from known malware. The number of malwares has expanded considerably in recent years, posing a serious security threat to financial institutions, businesses, and individuals. To stop malware from spreading, new methods for quickly recognizing and classifying malware samples so that their behaviour can be investigated are needed. In current Internet age, many virus attacks occur, posing serious security risks to financial institutions and everyday customers. The total number of malware occurrences has undoubtedly increased considerably over time. Here I use five machine learning classification model for the fast time in this dataset. I am classified according to the 54 correlated features with data visualizing, resizing and prepressing and finally proposed the best model for detection malware and model preparation method into many parts in this work. With almost 99% accuracy, the Random Forest Classifier outperforms. Second, with a score of 97 percent, K-Neighbors Classifier comes in second place in terms of malware classification accuracy. The rest of the models are less accurate.