Should I (re)Learn or Should I Go(on)?: Stream Machine Learning for Adaptive Defense against Network Attacks

Abstract

Continuous, dynamic and short-term learning is an effective learning strategy when operating in dynamic and adversarial environments, where concept drift constantly occurs and attacks rapidly change over time. In an on-line, stream learning model, data arrives as a stream of sequentially ordered samples, and older data is no longer available to revise earlier suboptimal modeling decisions as the fresh data arrives. Stream approaches work in a limited amount of time, and have the advantage to perform predictions at any point in time during the stream. We focus on a particularly challenging problem, that of continually learning detection models capable to recognize cyber-attacks and system intrusions in a highly dynamic and adversarial environment such as the open Internet. We consider adaptive learning algorithms for the analysis of continuously evolving network data streams, using (dynamic) sliding windows -- representing the system memory, to periodically re-learn, automatically adapting to concept drifts in the underlying data. By continuously learning and detecting concept drifts to adapt memory length, we show that adaptive learning algorithms can realize high detection accuracy of evolving network attacks over dynamic network data streams.