Applications of Machine Learning in Cyber Security - A Review and a Conceptual Framework for a University Setup

Abstract

Machine learning is a growing technical field due to its versatility and stability for the ever increasing data flow from heterogeneous sources and computational demands. Machine learning techniques are deployed nearly in every aspect of computing today because of its highly adaptive and scalable characteristics. It has the potential to adapt to new and unknown challenges. Cyber-Security is a field which is rapidly developing these days because of the attention that is required to secure the net-works and applications with the growth in social networks, internet and mobile banking, cloud computing, web technologies, smart grid etc. The domain of Cyber Security owing to its diversity and applications generate lot of data which is voluminous and coming from different sources. Such data provides great scope to the data scientists, cyber security specialists and machine learning enthusiasts, since they have the potential to provide insights which could help in curbing Cyber Crimes using Machine Learning Algorithms. This paper reviews literature in the field of machine learning and cyber security and highlighting the key developments and improvements in these fields. The applications of machine learning algorithms in cyber security have been discussed in detail in the paper. This paper focuses on the critical and the technical aspects of the previous work carried out by other researchers in these fields culminating with a comprehensive conclusion about the state-of-art in the fields. It also discusses as to what should be done to further improvise the situation currently faced by data scientists and cyber security researchers around the world. The current paper also discusses a conceptual framework for a typical university setup to tackle the Cyber Security issues and the implementation work on the proposed solution strategies is in progress and the authors are working on it.