|
Date: Wednesday 6/6/2012
Venue: MS020
Time: 1.00 pm
Speaker: Professor Hui Wang
Affiliation: School of Computing and Mathematics
University Of Ulster, Jordanstown
Graph Similarity -- Counting Through Matrices
By Professor Hui Wang
School of Computing and Mathematics
University Of Ulster, Jordanstown
Abstract:
Neighbourhood counting is a general methodology in designing combinatorial similarity, and is rooted in the concept of contextual probability. It has been specialised for different types of data, including multivariate, sequence and tree, resulting in different similarity measurements. In this talk, I will present recent work on graph similarity, which is inspired by the neighbourhood counting methodology.
I will first of all review the concepts of contextual probability and neighbourhood counting. I will then present a novel graph similarity, which is the number of all possible paths in a graph. An algorithm is devised to compute this similarity, which is based on adjacency matrix representation of graphs and is computed through matrix operations. This graph similarity is shown to be conceptually simple, mathematically concise, and practically tractable.
Short biography:
Hui Wang (BSc, MSc, DPhil) is Professor of Computer Science at the University of Ulster, Jordanstown Campus. His research interests include machine learning, data/text mining, uncertainty reasoning, and information retrieval. He has authored or co-authored 140+ journal/conference publications.
He is principal investigator of a number of regional, national and international projects (SAVASA, DEEPFLOW, BEACON, ICONS), and is co-investigator of several projects (STAR, TARSKI, UMIS). He is an associate editor of IEEE Transactions SMC-B, a member of the editorial board of International Journal of Computational Intelligence Systems, and an associate editor of International Journal of Machine Learning and Cybernetics.
He is chair of IEEE SMCS Ireland Chapter, and currently is a member of IEEE SMCS Board of Governors. |
