Author: Ashin KC
KC, Ashin, 2019 Algorithm for Cluster History Characterisation and Cluster Association: A graph based approach, Flinders University, College of Science and Engineering
Terms of Use: This electronic version is (or will be) made publicly available by Flinders University in accordance with its open access policy for student theses. Copyright in this thesis remains with the author. You may use this material for uses permitted under the Copyright Act 1968. If you are the owner of any included third party copyright material and/or you believe that any material has been made available without permission of the copyright owner please contact copyright@flinders.edu.au with the details.
The paper proposes new method for the cluster history characterization and finding cluster associations. Iterative searching for clusters over multiple stages in time span is relatively simple, however the knowledge of cluster behavior over time gives temporal characteristics of clusters which serves as a real utility on tracking the cluster evolution history. Similarity coefficient is used as major tool to quantify the cluster evolution path between two successive time intervals. A directed graph model is purposed for the representation of successive progression of cluster over time using similarity coefficient as the weighted edges between nodes. Searching and characterization process of evolving clusters of different instances or objects in a graph provides the frequent substructures and cluster path trajectory which is the significant step towards finding cluster associations.
Keywords: Data Mining, Higher Order Mining, Graph Mining, Cluster History Characterisation, Graph model, Similarity Coefficient
Subject: Engineering thesis
Thesis type: Masters
Completed: 2019
School: College of Science and Engineering
Supervisor: John Roddick