Protocol-based classification for intrusion detection

Authors:
Kun-Ming Yu;Ming-Feng Wu;Wai-Tak Wong
Affiliations:
Department of Computer Science and Information Engineering, Chung Hua University, HsinChu, Taiwan;Department of Information Management, Chung Hua University, HsinChu, Taiwan;Department of Information Management, Chung Hua University, HsinChu, Taiwan
Venue:
WSEAS Transactions on Computer Research
Year:
2008

Citing 2
Cited 1

Support-Vector Networks

Machine Learning
Learning intrusion detection: supervised or unsupervised?

ICIAP'05 Proceedings of the 13th international conference on Image Analysis and Processing

DDoS attacks detection model and its application

WSEAS Transactions on Computers

Quantified Score

Hi-index	0.00

Visualization

Abstract

A lightweight network intrusion detection system is more efficient and effective for real world requirements. Higher performance may result if insignificant and/or useless features are eliminated. Logistic Regression is one feature selection method. In this study, protocol type and Logistic Regression were used to pick up the feature sets which can get nearly the same performance as the full feature using a Support Vector Machine. Evaluation was done over a benchmark dataset used KDD CUP'99. In terms of time efficiency, the proposed method performs more than seven times better than other feature selection methods.