Information for Participants and Contributors
Organizer
Contact Us
- Email:
iccet@vip.163.com
- Tel: (International)
+86-28-8652-7699
- Tel: (中文)
4006-999-648
Relative Conferences
|
16-18 April 2010, Chengdu, China
Keynote Speakers
Keynote Speaker I

Chinese Academic of Engineering Fellow, Prof. Li Lemin
University of Electronic Science and Technology of China
中国工程院院士 电子科技大学 李乐民 教授
Research on Future Network Architecture
Abstract: The Internet has created great influence on our society, but there
are many problems, such as scalable, manageable, controllable, and
dependable, to be further solved. It is very important to study the future
of the Internet, and this is a hot topic in the world. In this talk, the
Chinese activity of research on future networks will be presented. The
technical problems related to future network architecture, such as
identifier and locator split, edge and core network split, and various
possible schemes will be discussed.
Keynote Speaker II

IEEE Fellow, Prof. Wang
Jun
The Chinese University of Hong Kong
- Professor, The Chinese
University of Hong Kong, Hong Kong, 2002 to present.
- Associate Professor,
The Chinese University of Hong Kong, Hong Kong, 1995 to
2001.
- Associate Professor,
University of North Dakota, Grand Forks, North Dakota,
USA, 1993-1997.
- Assistant Professor,
University of North Dakota, Grand Forks, North Dakota,
1990-1993.
- Associate Editor,
IEEE Transactions on Neural Networks, 1999 to
present.
- Associate Editor,
IEEE Transactions on Systems, Man, and Cybernetics -
Part B: Bybernetics, 2003 to present.
- Associate Editor,
IEEE Transactions on Systems, Man, and Cybernetics -
Part C: Applications and reviews, 2001-2005.
- Member of Editorial
Advisory Board, International Journal of Neural
Systems, 2006-2008.
- Guest Co-editor,
special issues in European Journal of Operational
Research (vol. 93, no. 2, 1996), International
Journal of Neural Systems (vol.
17, no. 6, 2007), and Neurocomputing (vol. 71,
nos. 16-18., 2008).
- Various positions at
numerous international conferences such as ISNN2004,
ISNN2005, IROS2005, ISNN2006, WCCI2006, ICONIP2006, and
WCCI2008.
- Past President of the
Asia Pacific Neural Network Assembly (APNNA).
Keynote Speaker III

IEEE and IAPR Fellow, Prof. B. John Oommen
School of Computer Science, Carleton University, Canada
Dr. John Oommen was born in Coonoor, India on September 9, 1953. He obtained
his B.Tech. degree from the Indian Institute of Technology, Madras, India in
1975. He obtained his M.E. from the Indian Institute of Science in
Bangalore, India in 1977. He then went on for his M.S. and Ph. D. which he
obtained from Purdue University, in West Lafayettte, Indiana in 1979 and
1982 respectively. He joined the School of Computer Science at Carleton
University in Ottawa, Canada, in the 1981-82 academic year. He is still at
Carleton and holds the rank of a Full Professor. Since July 2006, he has
been awarded the honorary rank of Chancellor's Professor, which is a
lifetime award from Carleton University. His research interests include
Automata Learning, Adaptive Data Structures, Statistical and Syntactic
Pattern Recognition, Stochastic Algorithms and Partitioning Algorithms. He
is the author of more than 325 refereed journal and conference publications,
and is a Fellow of the IEEE and a Fellow of the IAPR. Dr. Oommen has also
served on the Editorial Board of the IEEE Transactions on Systems, Man and
Cybernetics, and Pattern Recognition.
Keynote Talk: ICCET 2010
Title: Recent Advances in Learning Automata Systems
Abstract
This keynote talk will present the most recent advances in the theory and
applications of Learning Automata (LA) systems. Initially, it will
concentrate on the general area of stochastic LA, which are probabilistic
finite state machines that have been used to model how biological systems
can learn. The structure of such a machine can be fixed, or it can be
changing with time. We will explain how LA can also be implemented using
action (choosing) probability updating rules, which, in turn, may or may not
depend on estimates from the Environment being investigated.
While, traditionally, these updating rules have worked with the continuous
probability space, we will also describe how LA can be designed by
discretizing the probability space. The talk will describe the design and
analysis of both continuous and discretized LA, and will highlight the
subtle differences between the corresponding learning machines, their
convergence properties, and their learning capabilities.
The talk will then discuss the most recent developments such as the
Generalized Thathachar-Sastry estimator scheme.
The talk will include a comprehensive list of the applications in which LA
have proven their powerful potential.
Keynote Speaker IV

Dean Prof. Pan Wei
Southwest Jiaotong University
西南交通大学 信息科学与技术学院院长 潘炜 教授
Keynote Speaker V

Prof. Michal Wozniak, Ph.D., D.Sc.
Wrocsaw University of Technology,
Chair of Systems and Computer Networks, Faculty of Electronics,
Wybrzeze Wyspianskiego 27, 50-370 Wroclaw, Poland
Michal Wozniak is Professor of Computer
Science in the Department of Systems and Computer Networks, Faculty of
Electronics, Wroclaw University of Technology, Poland. He received an M.S.
degree in Biomedical Engineering in 1992 and Ph.D. degrees in Computer
Science in 1996 from the Wroclaw University of Technology. His research
focuses on multiple classifier systems, machine learning, data and web
mining, Bayes compound theory, distributed algorithms, computer and networks
security and teleinformatics.
Prof. Wozniak has published over 100 papers, 2 books and edited 3 books
Computer Recognition Systems (Springer). He is editor in chief of
International Journal of Computer Networks and Communications and
associate editor of several international journals including Pattern
Analysis and Applications, Expert Systems, Computational
Intelligence, and International Journal of Communication Networks and
Distributed Systems. He serves on program committees of numerous
international conferences. His works have been transitioned into commercial
applications.
See detail profile at:
http://www.kssk.pwr.wroc.pl/pracownicy/michal.wozniak-en
Combining pattern recognition algorithms – chances
and limits
Problem of
pattern recognition is accompanying our whole life. We start learn how to
recognize simple objects like “dog”, “flower”, “car” when we are young and
more sophisticated ones when we are growing up. Therefore methods of
automatic pattern recognition is one of the main trend in Artificial
Intelligence. The aim of such task is to classify the object to one of
predefined categories, on the basis of observation of its features. Such
methods are applied to many practical areas like prediction of customer
behavior, fraud detection, medical diagnosis etc. Numerous approaches have
been proposed to construct efficient, high quality classifiers like neural
networks, statistical and symbolic learning. There is much current research
into developing even more efficient and accurate recognition algorithms.
Multiple classifier systems are currently the focus of intense research. The
subject matter has been known for over 15 years. Some works in this field
were published as early as the ’60 of the XX century, when it was shown that
the common decision of independent classifiers is optimal, when chosen
weights are inversely proportional to errors made by the classifiers. In
many review articles this trend has been mentioned as one of the most
promising in the field of the pattern recognition. In the beginning in
literature one could find only majority vote, but in later works more
advanced methods of finding a common solution to the classifier group
problem were proposed. Estimation accuracy of the classifier committee is
one of fundamental importance. Known conclusions, derived on analytic way,
concern particular case of the majority vote when classifier committee is
formed on the basis of independent classifiers. Unfortunately this case has
only theoretical character and is not useful in practice. The weighted
voting is taken into consideration, but a problem of establishing weights
for mentioned voting procedure is not simple. Many of authors have proposed
treating the voting block as a kind of classifier but the general question
is “does the fuser need to be trained?”. An alternative way of common
classifier construction is combination of discriminants of available
classifiers in order to obtain set of common discriminants, e.g. via linear
combination.
The proposed
speech will present short review of the main methods of combined pattern
recognition, comparative analysis of some methods of classifier fusion based
on weighted voting of classifiers’ responses and combination of classifiers’
discriminants. The presented remarks will be illustrated by the results of
experimental based on real classification problems.
Keynote Speaker VI

Prof. Venkatesh Mahadevan,
Swinburne University of Technology, Australia
Dr Venkatesh Mahadevan is currently
lecturering in Information Systems (IS) and eBusiness the Faculty of
Higher Education at Lilydale campus of Swinburne University of
Technology, Australia. He has extensive research experience in the
different phases of the usability design lifecycle, from user research,
requirement specifications and conceptual design through to usability
attributes testing of a Telecollaboration (TC) business system.
Previously Venkatesh was attached to the Faculty of Engineering at the
University of Technology, Sydney (UTS) as Senior Research
Associate/Lecturer working for a CSIRO-led CeNTIE (Centre for Networking
Technologies for the Information Economy) TC project. He has a PhD from
UTS in which he built a models and application framework for a
Telecollaboration business system from the ground up in an agile
collaborative environment that would enhance the virtually enabled
collaborative practices of industry. This contemporary research from
several perspectives (such as business process, knowledge, service,
application and cognition) has met the necessity of addressing many of
the theoretical, practical and methodological issues surrounding
research on eBusiness/IT interface generally, and more specifically,
research into the value-added Human Computer Interaction (HCI). In
particular, he has researched how to integrate models from various
domains into an integrated cognitive system without leveraging the
usability attributes of a TC business system.
Venkatesh has presented his research
findings at several Australian and international conferences and
workshops. He has previously taught in India, Malaysia, and New Zealand.
During his stay at UTS he was also actively involved in teaching and
supervising undergraduate and postgraduate students. His personal
pedagogy improved greatly with a Graduate Certificate in Higher
Education Teaching and Learning conducted by Institute for Interactive
Media and Learning at UTS. Since then designing, developing,
implementing and publishing new pedagogic techniques are his favorite
academic activities.
Keynote Speaker VII

Prof. Zhou Tao
University of Electronic Science and Technology of China
电子科技大学 周涛 教授 Tao Zhou received
his Bachelor’s degree from the University of Science and Technology of
China in 2005. He obtained his Ph. D. majoring theoretical physics from
the University of Fribourg in 2010. In March 2010, he was awarded a full
professor position in the University of Electronic Science and
Technology of China. His main research interests include infophysics,
complex networks and human dynamics. He has published more than 110 SCI
journal papers, and about 50 papers have been published in prestigious
journals, including PNAS, Physical Reviews, New Journal of Physics and
Europhysics Letters. Reports on his works appeared in many mainstream
academic media, including Nature News, PNAS News, My Science, PhysOrg,
TG Daily, KijK, etc. All his publications have received more than 1200
citations according to the Web of Science, and his H-index is equal to
18.
Title: Solving the apparent diversity-accuracy dilemma of recommender
systems Abstract:
Recommender systems use data on past user preferences to predict
possible future likes and interests. A key challenge is that while the
most useful individual recommendations are to be found among diverse
niche objects, the most reliably accurate results are obtained by
methods that recommend objects based on user or object similarity. In
this presentation, we discuss how to comprehensively evaluate a
recommender system, introduce a new algorithm specifically to address
the challenge of diversity and show how it can be used to resolve this
apparent dilemma when combined in an elegant hybrid with an
accuracy-focused algorithm. By tuning the hybrid appropriately we are
able to obtain, without relying on any semantic or context-specific
information, simultaneous gains in both accuracy and diversity of
recommendations. Beyond the static algorithm, the dynamic environment in
the real world calls for adaptive techniques for information filtering,
namely to provide real-time responses to changing data. Many incremental
algorithms are designed for this aim, but they degrade from cumulative
errors over time. Here we give an incremental diffusion-based algorithm,
which integrate local and fast updating to achieve approximate results.
In addition to the fast responses, the errors of our proposed algorithms
do not cumulate over time, thus global recomputing is unnecessary.
Technique Papers:
T. Zhou, et al., Europhysics Letters 81 (2008) 58004;
T. Zhou, et al., New Journal of Physics 11 (2009) 123008;
T. Zhou, et al., PNAS 107 (2010) 4511.
|
Support organizations
|