Big Data Classification: Problems and Challenges in Network Intrusion Prediction with Machine Learning

UNCG Author/Contributor (non-UNCG co-authors, if there are any, appear on document)
Shanmugatha "Shan" Suthaharan, Associate Professor (Creator)
The University of North Carolina at Greensboro (UNCG )
Web Site:

Abstract: This paper focuses on the specific problem of Big Data classification of network intrusion traffic. It discusses the system challenges presented by the Big Data problems associated with network intrusion prediction. The prediction of a possible intrusion attack in a network requires continuous collection of traffic data and learning of their characteristics on the fly. The continuous collection of traffic data by the network leads to Big Data problems that are caused by the volume, variety and velocity properties of Big Data. The learning of the network characteristics requires machine learning techniques that capture global knowledge of the traffic patterns. The Big Data properties will lead to significant system challenges to implement machine learning frameworks. This paper discusses the problems and challenges in handling Big Data classification using geometric representation-learning techniques and the modern Big Data networking technologies. In particular this paper discusses the issues related to combining supervised learning techniques, representation-learning techniques, machine lifelong learning techniques and Big Data technologies (e.g. Hadoop, Hive and Cloud) for solving network traffic classification problems.

Additional Information

Performance Evaluation Review, 41(4), 70-73
Language: English
Date: 2014
Big Data, Hadoop distributed file systems, intrusion detection, machine learning

Email this document to