¹Software College, Northeastern University, Shenyang, China, 110819
²Department of Computer Science, Aberystwyth University, Aberystwyth, United Kingdom, SY23 3DB

---

Abstract—As network-based technologies become omnipresent, intrusion detection and prevention for these systems become increasingly important. This paper proposed a modified mutual information-based feature selection algorithm (MMIFS) for intrusion detection on the KDD Cup 99 dataset. The C4.5 classification method was used with this feature selection method. In comparison with dynamic mutual information feature selection algorithm (DMIFS), we can see that most performance aspects are improved. Furthermore, this paper shows the relationship between performance, efficiency and the number of features selected

Index Terms—Feature selection, classification, C4.5, intrusion detection, mutual information

[PDF]

Cite: Jingping Song, Zhiliang Zhu, Peter Scully, Chris Price, "Modified Mutual Information-based Feature Selection for Intrusion Detection Systems in Decision Tree Learning," Journal of Computers vol. 9, no. 7, pp. 1542-1546, 2014.