HIS'03 - Tutorials
Tutorials
will be held on Sunday, December 14, 2003.
Overview:
1. Information Assurance and Security
2. Intelligent Feature Extraction...
3. Innovative Soft Computing Applications for Mobile
Communication
4. Bayesian AI Tutorial
5. The Top 10 Data Mining Mistakes
6. The Brain and Mind Tissue...
1. Information
Assurance and Security
Author(s): Andrew H. Sung and Srinivas Mukkamala
Department of Computer Science
New Mexico Tech.
E-mail: sung@cs.nmt.edu, srinivas@cs.nmt.edu
Abstract. As a result of the rapidly increasing incidents of security
breaches and malicious attacks, and the heightened concern for cyber terrorism,
there is an increasing need for governments,
organizations, enterprises, and individuals to employ enhanced security
measures and security
devices to protect their computer systems and information assets.
This tutorial begins with an introduction to the basic concepts and issues
of information
assurance. An assortment of important current topics will be discussed
next, including
vulnerability analysis, computer attacks (with in-depth coverage of denial
of service attacks),
intrusion detection, and software security assurance. Current research
in selected areas will also
be presented to give the audience an understanding of the technical challenges
involved and the
techniques being explored in information security.
About the Authors.
Andrew H. Sung is currently Professor and Chairman of the Computer
Science Department of
New Mexico Tech, and a founding coordinator of the schools new Information
Technology
Program. He is also the Associate Director for Education and Training
of ICASA (Institute for
Complex Additive Systems Analysis, a research division of New Mexico Tech
performing work
on information technology, information assurance, and analysis and protection
of critical
infrastructures as complex interdependent systems).
Andrew Sung received his Ph.D. in Computer Science from the State University
of New York at
Stony Brook in 1984. He joined New Mexico Tech in 1987, and served as
the Computer Science
department chair from 1988 to 1993, and again since January 2000. He has
over 100 publications
and his current research focus is information security.
Srinivas Mukkamala is currently a doctoral student of the
Computer Science Department of New
Mexico Tech. He is currently working in the areas of information assurance
and security and has
over 30 publications in the areas of information security.
Srinivas Mukkamala received his B.E. in Computer Science and Engineering
from University of
Madras in 1999, M.S. in Computer Science form New Mexico Tech. He is currently
a research
assistant and a student lead of the information assurance research group
at New Mexico Tech.
2. Intelligent
Feature Extraction from Knowledge for Forecasting and Decision Making
Author(s): Parag Kulkarni
Scientific Applications Center
Siemens Information Systems Ltd.
Shivaji Nagar, Pune 411 016
India
Tel.: +91-20-565-1744 Extn. 154
Fax: +91-20-400-2459
E-mail: parag.kulkarni@sisl.co.in, parag.kulkarni@siemens.com
Abstract. Decision system is key are in business and knowledge
industry. Better feature
extraction and organized methodologies can lead to better decision making.
Many
times user feels that domains of application for it are limited and those
are not useful
for their business solution. Here we want to clarify it can be used very
effectively in
all domains. It can help for decision making, better management. There
are many
different methodologies those can be developed using basic methodologies
and used
in numerous applications. Purpose of this tutorial is to give an overview
of
application domains, introduce open areas of research in this field and
present
effectiveness of decision systems with case studies. This tutorial ill
be useful for
students and professionals working in the area of decision systems or
are active in
optimization of one of the domains related to it.
About the Author. Ph.D. from IIT Kharagpur, Working in IT industry
for more than 12 years.
Areas of interest: Decision systems, forecasting, distributed computing,
AI, expert
systems.
3.Innovative
Soft Computing Applications for Mobile Communication
Author(s): Suthikshn Kumar
Communication and Embedded Systems
Larsen & Toubro Infotech Limited
4th Floor, #2 Church Street
Bangalore 560 001
India
Tel.: +91-80-5323734/5/6
Fax: +91-80-5323738
Mobile: +91-9844106248
E-mail: suthikshn.kumar@lntinfotech.com
Web: http://www.lntinfotech.com, http://www.larsentoubro.com
Abstract. Soft Computing techniques comprise of topics from Fuzzy
logic, Neuro computing, evolutionary computing, probabilistic computing,
chaotic computing and machine learning. They are frequently being used
where the mathematical model for given problem is not available. It is
a step in lateral thinking for solving numerous unsolved problems. As
the use of soft computing leads to High Machine IQ ( HMIQ), the resulting
smart systems find innovative applications.
The soft computing techniques are increasingly being used for mobile communications.
In this session, the speakers explore the design and development of some
innovative applications of soft computing for mobile communication.
- Smart Volume Tuner for Cellular Phones: Using the information on background
noise levels, a fuzzy logic controller adjusts the acoustic volume levels
of mobile handset. This leads to a smart mobile phones which delivers
an improved speech quality.
- Voice dialling for mobile phones: A genetic-fuzzy algorithm used for voice
dialling application. The resulting smart dialling application is described.
- Background Noise classification and Voice Activity Detector using soft
computing techniques
The demo of these applications will be provided at the end of the session.
Also, we will distribute demo CDs to the attendees with the product brochures.
About the Author. Dr. Suthikshn Kumar is currently Project Manager
in the Communications and Embedded Systems Division at Larsen & Toubro
Infotech Limited. He studied engineering at Bangalore University, Indian
Institute of Technology where he received his BE and MTech Degrees respectively.
He received the PhD degree from the University of Melborne(Australia).
Dr. Kumar is a Senior Member of IEEE, Member of IETE, CSI, ACM, SPIN,
YRC-WFSC. He is the founding chair of Bangalore chapter of ACM. He has
received employee recognition awards at Philips Semiconductors and InfineonTechnologies
3i awards for innovation. He has published several papers at major international
conferences/conventions/journals. He is the chair of HiPC workshop on
Soft Computing ( WoSCo'02 and WoSCo03).
4. Bayesian AI
Tutorial
Author(s): Kevin B. Korb and Ann E. Nicholson
School of Computer Science and Software Engineering
Monash University, Victoria 3800
Australia
Tel.: +3-9905-5198/5211
Fax: +3-9905-5146
E-mail: korb@csse.monach.edu.au, annn@csse.monach.edu.au
Web: http://csse.monash.edu.au/bai/
Abstract. Bayesian networks have rapidly become one of the leading
technologies for
reasoning under uncertainty, by explicitly modeling causal relationships
as well as
supporting decision analysis. This follows the work of Pearl, Lauritzen,
and others
in the late 1980s using graphical models to make Bayesian reasoning feasible.
In this tutorial, we begin with a brief examination of the philosophy
of Bayesian-
ism, motivating the use of probabilities in decision making, agent modeling
and
data analysis. We introduce Bayesian networks for modeling and reasoning
under
uncertainty and provide an overview of the inference techniques, including
causal
inference. We then describe two extensions of Bayesian networks: (1) decision
networks, which explicitly support decision making under uncertainty,
and (2) dy-
namic Bayesian networks, which allow explicit reasoning about changes
over time.
We will illustrate the presentation throughout with examples using the
Netica
Bayesian network software.
The biggest obstacle to Bayesian AI having a broad and deep impact outside
of
the research community are the dfficulties in developing applications,
difficulties
with eliciting knowledge from experts, and integrating and validating
the results.
This has led to the rapid growth in applying machine learning methods
to the
automated building of Bayesian networks. In this tutorial, we review the
main
data mining techniques available for learning Bayesian network structures
and for
parameterizing them, including dealing with incomplete data.
Another issue is that there is as yet no clear methodology for developing,
testing and deploying Bayesian network technology in industry and government
-
there is no recognized discipline of "software engineering"
for Bayesian networks.
In this tutorial we introduce the process of Knowledge Engineering with
Bayesian
Networks (KEBN) together with some tools we have developed in support.
We conclude the presentation part of the tutorial with two case studies
of
Bayesian network development: the first an ecological risk assessment
model de-
veloped in the Monash Water Studies centre, and the second SARBayes, a
collab-
orative project with Victorian Search and Rescue. Following the formal
tutorial
presentation in the morning, there will be an afternoon lab session where
partici-
pants will undertake some modeling exercises using Netica.
About the Authors.
Ann Nicholson completed her doctorate in the robotics research
group at Ox-
ford University (1992) working on dynamic belief networks for discrete
monitoring.
She then spent two years at Brown University as a postdoctoral research
fellow
before taking up a lecturing position at Monash University in Computer
Science.
Her general research focus is AI methods for reasoning under uncertainty,
while
her current research areas include approximate methods for Bayesian networks,
evaluation, applications of dynamic belief networks and approximate planning
us
ing Markov Decision Processes. She has taught numerous subjects in Computer
Science, including Artificial Intelligence, and developed and taught a
webbased
introduction to Lisp programming.
Kevin Korb did his PhD in the philosophy of science at Indiana
University
(1992) working on the philosophical foundations for the automation of
Bayesian
reasoning. Since then he has lectured at Monash University in Computer
Science,
combining his interests in philosophy of science and artificial intelligence
in work
on understanding and automating inductive inference, the use of MML in
learning
causal theories, artificial evolution of cognitive and social behavior,
and modeling
Bayesian and human reasoning in the automation of argumentation. He has
in
dependently developed and taught the following subjects: Machine Learning
(3rd
year and honours); Bayesian Reasoning (honours); Causal Reasoning (honours);
The Computer Industry: Historical, Social and Professional Issues (3rd
year).
5. The Top
10 Data Mining Mistakes
Author(s): John F. Elder IV
Elder Research, Inc.
635 Berkmar Circle
Charlottesville, VA 22901
E-mail: elder@datamininglab.com
Tel.: (434)-973-7673
Fax: (703)-995-0387
Web: http://www.datamininglab.com
Abstract. The tutorial will reveal the top mistakes we Data Miners
can make, from the simple to the
subtle, using real-world (often humorous) stories. The topics will be
presented from case
studies of real projects and the (often overlooked) symptoms that suggested
something
might be amiss...
The goal will be to learn "best practices" from their flip side
-- mistakes. But also,
following the introduction of a topic (e.g., bootstrapping) the 3-hour
tutorial format will
allow for brief summaries of how to do it right -- that is, mini-tutorials
on the key
principles to keep in mind when using a particular Data Mining technique.
Mistakes to be covered include: Lack data, Focus on Training, Rely on
1 technique, Ask
the wrong question, Listen (only) to the data, Accept leaks from the future,
Discount
pesky cases, Extrapolate (practically and theoretically), Answer every
inquiry, Sample
without care, Believe the best model.
About the Author. Dr. John Elder heads a small Data Mining firm
with offices in Charlottesville, Virginia, and Washington, DC. John earned
degrees in Electrical Engineering at Rice University,
then worked in the Defense consulting industry for 5 years, where he authored
an early
Data Mining tool for the Air Force which led to improved guidance and
flight control
applications. He then earned a Ph.D. in Systems Engineering from the University
of
Virginia while working as Director of Research for an investment management
firm, and
wrote an influential tool for global optimization. After two years post-doctoral
research at
Rice in the Computational and Applied Mathematics Department, John returned
to
Virginia and started Elder Research, Inc. in 1995, where hes led
projects successfully
applying Data Mining to a wide variety of financial, commercial, and medical
applications -- including cross-selling, customer segmentation, direct
marketing, credit
scoring, sales forecasting, stock selection, drug efficacy, biometrics,
market timing, and
fraud detection. Dr. Elder has written several book chapters and articles
on pattern
discovery techniques, and is a frequently invited conference speaker.
He is active on
Statistical and Engineering journals and boards, and his popular Data
Mining courses are
acclaimed for clarity. He has been named to Whos Who in the World
for his
contributions to the field. Dr. Elder has been honored, since Fall 2001,
to serve on Panel
formed by Congress to guide critical defense technology for the National
Security
Agency.
6. The Brain
and Mind Tissue and Networks that Merge the Understanding, Consciousness,
Emotions and Knowledge
Author(s): Branko Soucek
E-mail: branko.soucek@libero.it
Abstract. Nested, Fractal, Time I Information I Space S set, TISS
is presented. TISS fine and course computing networks directly emulate
some of the functions of animal and human brain and mind. TISS net offers
a high level of the generalization, recognition, learning and converging.
TISS net chaotic selforganization leads to the new real - time applications:
in medicine, business, communication, industry, internet.
About the Author. Branko Soucek, professor at Universities of
Zagreb, New York and Arizona. Researcher and consultant for the United
Nation Agencies UNIDO and IAEA, and for NASA, IBM, Siemens, Schering,
Brookhaven National laboratory.
Prof. Soucek has pubblished 10 books with Wiley Inc. New York. His books
have been translated into the Croatian, Russian and Japanese languages.
He is a member of the Croatia Academy of Science and of the American Association
for Advancement of Science.
|