Projects per year
Abstract
In this work, a new method for classification is proposed consisting of a combination of feature selection, normalization, fuzzy C means clustering algorithm and C4.5 decision tree algorithm. The aim of this method is to improve the performance of the classifier by using selected features. The fuzzy C means clustering method is used to partition the training instances into clusters. On each cluster, we build a decision tree using C4.5 algorithm. Experiments on the KDD CUP 99 data set shows that our proposed method in detecting intrusion achieves better performance while reducing the relevant features by more than 80%.
Original language | English |
---|---|
Title of host publication | Cyberspace Safety and Security |
Subtitle of host publication | Proceedings 5th International Symposium, CSS 2013, Zhangjiajie, China, November 13-15, 2013 |
Editors | Guojun Wang |
Publisher | Springer Nature |
Pages | 299-307 |
Volume | 8300 |
ISBN (Electronic) | 978-3-319-03584-0 |
ISBN (Print) | 978-3-319-03583-3, 3319035835 |
Publication status | Published - 11 Nov 2013 |
Publication series
Name | Lecture Notes in Computer Science / Security and Cryptology |
---|---|
Publisher | Springer |
Keywords
- Intrusion detection
- Fuzzy C-Means
- Feature selection
- C4.5
Fingerprint
Dive into the research topics of 'Selecting Features for Anomaly Intrusion Detection: A Novel Method using Fuzzy C Means and Decision Tree Classification'. Together they form a unique fingerprint.Projects
- 1 Finished
-
Self-heating architectures against malware propagation
Scully, P. M. D.
01 Oct 2011 → 30 Sept 2014
Project: Externally funded research