Effective Malware Detection using Gradient Boosting and Convolutional Neural Network

Effective malware detection is the need of the hour for all industries. With the rise of intelligent polymorphic malware, it becomes a necessity to invest in developing systems to identify the malware and quarantine them. This paper tries to solve the problem of malware detection using static analysis and machine learning. We try to take a step forward in solving the problem. After going through multiple techniques used like heuristics, signature checking, machine learning models like multilayer perceptron (MLP), Random Forest (RF) and neural networks, we propose a 2-step solution that involves a binary classification model created using Gradient Boosting Classifier and the EMBER 2018 dataset and family classification model using Convolutional Neural Networks (CNN) and Virus MNIST dataset.