Managing Large-Scale Heterogeneous Deployments for Cybersecurity

Abstract

In cyber defense, we must contend with the massive amounts of data being generated in a variety of different formats and speeds. Unfortunately, traditional tools and methods are not meeting the requirements for scale and speed and rely too heavily on heuristics. Advancements in mobile technologies and the Internet of Things (IoTs) will continue to contribute to the additional growth in data volumes anticipated for the foreseeable future. As data continues to grow in complexity and scale, cyber professionals must rely upon models that are more elaborate and sophisticated to predict future behavior. More complex models can give additional inference capabilities; however, they are also difficult to scale and deploy in real-time environments. Managing large-scale, heterogeneous deployments for cybersecurity is challenging. Hardware capabilities and software tools both motivate and limit computational and inferential objectives. Hence, the interplay between data science (especially machine learning) and computation become more significant than ever to explore to gain more insight into heterogeneous deployments and how they can be more effectively managed. In this study, we identify ways in which data science tools and techniques can be used in improving the management of large-scale heterogeneous deployments for cybersecurity.