Home > Research > Publications & Outputs > Classification of Decision-Behavior Patterns in...
View graph of relations

Classification of Decision-Behavior Patterns in Multivariate Computer Log Data Using Independent Component Analysis

Research output: Contribution in Book/Report/Proceedings - With ISBN/ISSNConference contribution/Paperpeer-review

Published
Close
Publication date2003
Host publicationKnowledge-Based Intelligent Information and Engineering Systems 7th International Conference, KES 2003, Oxford, UK, September 2003. Proceedings, Part II
EditorsVasile Palade, Robert Howlett, Lakhmi Jain
Place of PublicationBerlin
PublisherSpringer
Pages73-79
Number of pages7
Volume2774
ISBN (print)978-3-540-40804-8
<mark>Original language</mark>English

Publication series

NameLecture Notes in Computer Science
PublisherSpringer Berlin / Heidelberg
Volume2774
ISSN (Print)0302-9743
ISSN (electronic)1611-3349

Abstract

The purpose of the work is to demonstrate the usefulness and applicability of audit trail data for modeling, classifying and predicting human user decision behavior using Independent Component Analysis. A hybrid behavior-modeling method, based on independent component analysis and sensitivity analysis, is proposed to directly classify decision behavior in terms of predetermined target measures. A comparison with alternative principal component analysis schemes demonstrates the advantages of the proposed method in terms of algorithmic speed, complexity, practical feasibility and data reduction capability. The capability of the system to provide intelligent support to executive decision making and internal control and to assist in several phases of the decision making process is highlighted.