Title: A rough set-based effective rule generation method for classification with an application in intrusion detection
Authors: Prasanta Gogoi; Dhruba K. Bhattacharyya; Jugal K. Kalita
Addresses: Department of Computer Science and Engineering, Tezpur University, Napaam, Tezpur 784028, Assam, India ' Department of Computer Science and Engineering, Tezpur University, Napaam, Tezpur 784028, Assam, India ' Department of Computer Science, College of Engineering and Applied Science, University of Colorado, Colorado Springs, CO 80918, USA
Abstract: In this paper, we use Rough Set Theory (RST) to address the important problem of generating decision rules for data mining. In particular, we propose a rough set-based approach to mine rules from inconsistent data. It computes the lower and upper approximations for each concept, and then builds concise classification rules for each concept satisfying required classification accuracy. Estimating lower and upper approximations substantially reduces the computational complexity of the algorithm. We use UCI ML Repository data sets to test and validate the approach. We also use our approach on network intrusion data sets captured using our local network from network flows. The results show that our approach produces effective and minimal rules and provides satisfactory accuracy.
Keywords: rough sets; LEM2; inconsistent data; minimal rules; redundant rules; PCS; intrusion detection; network flow data; decision rules; data mining; classification accuracy; network intrusion; network security.
International Journal of Security and Networks, 2013 Vol.8 No.2, pp.61 - 71
Available online: 16 Aug 2013 *Full-text access for editors Access for subscribers Purchase this article Comment on this article