A Review on Deep Learning Approaches to Real Time Network Intrusion Detection System
Organizations are adapting the network security technologies to protect their data and infrastructure in the wake of rapid increase in modern sophisticated cyber-attacks. IDS is one such system used by organization to differentiate between abnormal and normal behavior and identify the system attacks. But intrusion analysis of large data such as audit or log files with present IDS solution is not optimal since these IDS generate high False-Positive rate and Time to identify attack is also more. In recent times we have seen Machine learning being used to aid IDS to improve its performance. Machine learning based IDS identify sophisticated attack with better accuracy, reduce the False-Positive rate and in a timely manner. In this paper we will be looking at the different phases of an cyber-attack, and we will look at different Machine Learning algorithm such as Cluster-Based Approach, machine-Learning Based Approach, Optimization Algorithm based approach, artificial neural network based approach, deep learning based approach, model based approach, hybrid mock-up based approach. Here we will have a brief discussion on working of this algorithm, their shortcoming and their false positive accuracy rate with respect to each other.