Network Intrusion Detection Using Deep Learning

Abstract

Network intrusion detection is one among the foremost important parts for cyber security to guard computer systems against malicious attacks. With the emergence of various sophisticated and new attacks, however, network intrusion detection techniques face several significant challenges. We propose a completely unique network intrusion model by stacking autoencoders and evaluate our method on intrusion detection datasets. The auto-encoder is one among the foremost interesting models to extract features from the high-dimensional data within the context of deep learning. Our proposed model provides an accuracy which is more efficient than machine learning techniques like Random forest and Naive Bayes.