Network Intrusion Detection System Using Ensemble Methods and Deep Neural Network

Today, we find almost everybody using internet and with increase in demand of computer networking, Hackers are taking advantage of this situation and trying to intrude into the networks and disturb the networks thus by injecting malwares into it. It is very important to identify whether the network is free from malware or not in order to ensure secrecy of valuable data. Intrusion Detection Systems (IDS) are some of the most prominent technologies for administrating and looking after security issues in the network. With rapidly changing network traffic data, one classifier is not enough to fend off modern network intruders. In this work, ensemble classifiers are used to classify data based on binary and multiclass. We took this as a major challenge to increase accuracy rates of detection for individual attack types and all types of attacks. This allows us to identify attacks and specific attack categories. We've chosen to contrast other works by opting for Artificial Networks as well. Python being the most versatile programming language in recent times, we chose IDS-ML an open source code repository in order to develop IDS from public network traffic records that can protect modern networks with high efficiency and accuracy.