Network Security Threats Detection Methods Based on Machine Learning Techniques

The world is now interconnected to share information and resources in a significant manner. The network must support its integrity and network security, and it must be scalable to sustain its performance and capacity. The network system is subjected to attack and due to the greater extent of applications, development of technology, added users and IoT (Internet of Things), the network system is vulnerable to different security threats. The network security threats are framed and launched with highly effective techniques targeting the network resources. The computer network system must be protected against any attack to save its resource. Machine Learning techniques are preferably being used by researchers as a solution to these problems. The development of machine learning shows effective capabilities to analyse and detect, sometimes also to prevent network security threats. In the last few years, various machine learning algorithms and applications have been designed and experimented with network traffic analysis, including Data Mining and Network-based techniques in the field of Data Science. In this work, we have conducted a detailed analysis of various machine learning techniques. Some Machine Learning approaches and related algorithms are reviewed in connection with their applications to analyse and detect network threats and the prospective solutions proposed in various research works.