Advance Deep Learning Technique for Big Data Classification in IDS Environment

Abstract

Deep-learning techniques are utilized extensively to construct an intrusion detection system (IDS) for the timely and automated detection as well as classification of cyber assaults at network and host levels. Many difficulties exist, however, because harmful attacks continue to change and require a scalable solution in very high numbers. Various IDS big datasets are freely available by the cyber security community for future investigation. However, no current work has shown an exhaustive evaluation the malware data sets made available to the public must be consistently updated and benchmarked. The construction of a flexible and efficiently Hybrid FFNN, a kind of deep learning model, to recognize and classify unforeseen and unplanned cyber-attacks is discussed in this document.