Multi-Layered Filtration Framework for Efficient Detection of Network Attacks Using Machine Learning

The advancements and reliance on digital data necessitates dependence information technology. growing amount of their availability over the Internet have given rise to problem security. With increase in connectivity among devices networks, maintaining security an asset has now become essential for organization. Intrusion detection systems (IDS) are widely used networks protection against different network attacks. Several machine-learning-based techniques been researchers implementation anomaly-based IDS (AIDS). In past, focus primarily remained improvement accuracy system. Efficiency with respect time is important aspect IDS, which most research thus far somewhat overlooked. For this purpose, we propose a multi-layered filtration framework (MLFF) feature reduction using statistical approach. proposed helps reduce without affecting accuracy. We use CIC-IDS2017 dataset experiments. contains three filters connected sequential order. accuracy, precision, recall F1 score calculated selected machine learning models. addition, training also because these parameters considered measuring performance Generally, decision tree models, random forest methods, artificial neural show better results attacks minimum time.