|
|
|
|
Mail: |
Dr. L.I.
Kuncheva, |
|
Phone: |
+44 (0) 1248
383661 |
|
Email: |
Biography
I graduated in Electrical Engineering from the
Technical University of Sofia, Bulgaria in 1982, and then took up a 2-year
postgraduate course in Applied Mathematics. I obtained my PhD in 1987 from the
Research interests
My
research has always gravitated towards pattern recognition and classification.
I have dabbled in almost all aspects of the classical paradigm: feature
selection, classification, clustering, data editing, etc.
At the
start I was intrigued by the possibility of designing simple, accurate and interpretable
classifiers using the advances of fuzzy set theory (Fuzzy
Classifier Design, Springer, 2000.) Maybe the fascination came from the
fervent arguments between proponents and opponents, and the natural human
impulse to side with the contestant being attacked. I used to serve as an
Associate Editor for IEEE Transactions on
Fuzzy Systems from 1997 to 2001, and was on the Editorial board of Fuzzy Sets and Systems between 1998 and
2003. I recently received an award for Outstanding Paper in the IEEE
Transactions on Fuzzy Systems for my paper “Fuzzy vs Non-fuzzy in
combining classifiers designed by boosting”, IEEE Transactions on Fuzzy Systems, 11 (6), 2003, 729-741. It was
fun while it lasted.
Since
about 1990 I have been working in classifier combination (Combining
Pattern Classifiers. Methods and Algorithms, Wiley, 2004). Various classifier
combination streams existed in disguise before Fabio Roli and Josef Kittler set
up in year 2000 a regular pattern of Workshops on Multiple Classifier Systems,
unifying the terminology and bringing together machine learning and pattern
recognition communities. My main topic of interest was diversity in classifier
ensembles and how it will help to understand and improve classifier ensembles.
It turned out that the relationship between diversity and ensemble performance
was not as straightforward as one would expect, so the challenge took the
fantasy of other researchers. I was a guest editor of a special issue of
Information Fusion on diversity in classifier ensembles (Volume
6, Issue 1, Pages 1-115, March 2005, Diversity in Multiple Classifier Systems)
My
current research interests are broadened by new projects and collaborations. I
am leading a project on classifier ensembles for changing environments (EPSRC).
I am supervising a PhD on image analysis for micro-palaeontology (CASE EPSRC)
and another in feature selection for diagnosing Scrapie in sheep and BSE in
cattle. Visitors
to our group share my interests in classifier ensembles and cluster ensembles,
and I take great pleasure in working with every one of them.
Books
|
|
|
|
|
|
Combining
Pattern Classifiers. Methods and Algorithms, Wiley, 2004.
|
|
Fuzzy
Classifier Design, Springer, 2000.
|
Matlab code
(selected)
|
|
|
|
|
|
|
||
Teaching
|
Signature |
Module
title |
|
ICP1024 |
Algorithm
Design (06/07 - present ) |
|
IPS3083 |
Pattern Recognition and Neural Networks (04/05 - present) Statistical
Pattern Recognition (98/99 - 03/04) |
|
IOR3086 |
Heuristic Optimization (00/01 -
present) - once in 2 years |
|
IOR2085 |
Operational Research 1 (01/02 -
present) |
|
IPR3084 |
Fuzzy Sets, Uncertainty and Information
(98/99 - 00/01, 06/07) |
|
IPS1080 |
Probability
and Statistics (01/02 - 04/06) Probability
Theory (98/99 - 00/01) |
|
IPS2005 |
Data
Analysis (02/03 - 04/05) |
|
G2M52 |
Applications
of Mathematics (99/00 - 00/01) |
Return to
Bangor's Home Page
Return
to Mathematics
Home Page
Last
modified: 13/09/06