Design of a hierarchical principal component analysis system for field intrusion detection

Aleksey Y. Ashikhmin, James H. Graham, Ahmed H. Desoky, Benjamin Arazi

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Scopus citations

Abstract

Hierarchical Principal Component Analysis (HPCA) is a fast data reduction technique for Field Intrusion Detection System (FIDS). This paper presents a mathematical model for group assignment of HPCA. For testing purposes a chemical distillation process controlled by SCADA system was used to perform real-time analysis and evaluation of the proposed FIDS. Detailed analysis of the FIDS is discussed including the receiver operating characteristic curve (ROC), clustering variables, implementation of a low-pass filter to reduce high noise, and the sensitivity of HPCA versus PCA. Experimental intrusion detection results are promising.

Original languageEnglish
Title of host publication22nd International Conference on Computer Applications in Industry and Engineering 2009, CAINE 2009
Pages125-132
Number of pages8
StatePublished - 1 Dec 2009
Externally publishedYes
Event22nd International Conference on Computer Applications in Industry and Engineering 2009, CAINE 2009 - San Francisco, CA, United States
Duration: 4 Nov 20096 Nov 2009

Publication series

Name22nd International Conference on Computer Applications in Industry and Engineering 2009, CAINE 2009

Conference

Conference22nd International Conference on Computer Applications in Industry and Engineering 2009, CAINE 2009
Country/TerritoryUnited States
CitySan Francisco, CA
Period4/11/096/11/09

Keywords

  • Computer security
  • Network intrusion detection
  • Principal components analysis
  • SCADA

Fingerprint

Dive into the research topics of 'Design of a hierarchical principal component analysis system for field intrusion detection'. Together they form a unique fingerprint.

Cite this