Machine Learning Based Intrusion Detection System for Network Security Using Self-Organizing Map

Abstract

Machine learning techniques are widely used for detecting network attacks at the network level and the host level in a timely and automatic, to develop an intrusion detection system (IDS) for grading. Malicious attacks are always the need for change and scaling solutions, because it occurred during a very large volume, however, a number of challenges will occur. There is a data set that can be publicly available for further research on another malware of cyber security within the community. Network, plays an important role in modern life, has become the network security is an important field of research. Is an important network security technology Intrusion Detection System (IDS) being to monitor the software running on the network and hardware status. Despite the decades of development, the existing IDS, still, to reduce the improved error rate detection accuracy, are faced with the challenge of detecting unknown attacks. To overcome the issues proposed the method is Self-Organizing Map (SOM) Performance of intrusion detection depends mainly on the accuracy. Accuracy intrusion detection is to reduce the error rate, it must be strengthened in order to increase the detection rate. In order to improve performance, different technologies, have been used in recent works. It is the main work of the intrusion detection system for analyzing a huge network traffic data. Well-organized classification method, you need to solve this problem.