A Hybrid Feature Selection Approach based on Random Forest and Particle Swarm Optimization for IoT Network Traffic Analysis

The complexity and volume of network traffic has increased significantly due to the emergence “Internet Things” (IoT). classification accuracy is dependent on most pertinent features. In this paper, we present a hybrid feature selection method that takes into account optimization Particle Swarms (PSO) Random Forests. data collected by security firm, CIC-IDS2017, contains large number attacks instances. To improve accuracy, use framework's RF algorithm identify important Then, PSO used refine process. According our experiments, proposed performed better than other methods when it comes accuracy. It achieves ~99.9% using Forest PSO. approach also helps model's performance. suggested can be utilized analysts administrators prevent IoT.