Large-scale nonlinear dimensionality reduction for network intrusion detection

Abstract : Network intrusion detection (NID) is a complex classification problem. In this paper, we combine classification with recent and scalable nonlinear dimensionality reduction (NLDR) methods. Classification and DR are not necessarily adversarial, provided adequate cluster magnification occurring in NLDR methods like $t$-SNE: DR mitigates the curse of dimensionality, while cluster magnification can maintain class separability. We demonstrate experimentally the effectiveness of the approach by analyzing and comparing results on the big KDD99 dataset, using both NLDR quality assessment and classification rate for SVMs and random forests. Since data involves features of mixed types (numerical and categorical), the use of Gower's similarity coefficient as metric further improves the results over the classical similarity metric.
Complete list of metadatas

Cited literature [19 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-01517215
Contributor : Ludovic Journaux <>
Submitted on : Tuesday, August 27, 2019 - 4:48:43 PM
Last modification on : Wednesday, September 18, 2019 - 4:04:05 PM

File

Hamid2017.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01517215, version 1

Citation

Yasir Hamid, Ludovic Journaux, John Aldo Lee, Lucile Sautot, Bushra Nabi, et al.. Large-scale nonlinear dimensionality reduction for network intrusion detection. 25th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN 2017), Apr 2017, Bruges, Belgium. pp.153-158. ⟨hal-01517215⟩

Share

Metrics

Record views

297

Files downloads

5