题 目：Smart Network Defense through Perpetual Learning
Perpetual learning, also known as continuous, life-long, or never-ending learning, an active research direction in machine learning, is concerned with how to develop computing systems that can automatically, consistently and continuously improve their performance at tasks over time. How to safeguard network infrastructure and computing resources for an enterprise in the presence of ever-increasing attacks and exploitations has become a very challenging issue. A critical capability in a network defense system is how to plug the holes in existing security polices and adapt to emerging and new threats and attacks so as to incrementally improve its defensive prowess over time. We believe that perpetual learning offers a viable approach toward building such a smart network defense.
In this talk, we use Snort, a network intrusion detection and prevention system, as the backdrop, and discuss how perpetual learning can be utilized to augment the network defense system with the capability of incremental performance improvement. Learning episodes to engage in incremental performance improvement will be triggered by encountered deficient phenomena. Heuristics specific to the deficient phenomenon will be deployed in each learning episode. The outcome of such a learning episode is a refined or augmented knowledge base that codifies improved security policies. The performance improvements are essentially embodied in security holes being plugged or new threats becoming identifiable. Perpetual learning thus enables a network defense system to become smart over an open-ended sequence of many learning episodes.
Du Zhangis Professor and Dean of the Faculty of Information Technology, Macau University of Science and Technology, Macau, China. He received his Ph.D. and M.S. degrees, both in computer science, from the University of Illinois and Nanjing University, China, respectively. Previously he was a Professor and Chair of the Computer Science Department at California State University, Sacramento. He has research affiliations with numerous universities in the USA, UK, France, Hong Kong, Czech Republic, and Mexico. Professor Zhang's current research interests include machine learning (inconsistency-induced perpetual learning), knowledge-based systems, big data analytics, and software engineering. He has over 200 publications in these and other areas. He has served in various roles on numerous international conferences, and is editor or editorial board member for several journals in the areas of artificial intelligence, software engineering and knowledge engineering, big data, and applied mathematics.