|Sami Khuri's Colloquia Presentations|
Bioninformatics Department of Mathematics, SJSU, October 2002.
Abstract: The rapidly emerging field of bioinformatics promises to lead to advances in understanding basic biological processes, and in turn, advances in the diagnosis, treatment, and prevention of many genetic diseases. After a very brief refresher on molecular biology, namely what is known as the central dogma of molecular biology, the presentation will move to bioinformatics. We first define what is meant by Bioinformatics and then move to identifying some of the challenging problems encountered in that field. The challenges in which computer scientists and mathematicians can play a role include, pairwise and multiple sequence alignments, database searches, phylogenetic analysis, gene prediction, structure prediction, and fragment assembly.
Genetic Algorithms Department of Biological Sciences, SJSU,
Genetic Algorithms short course, Department of
Computer Science and Engineering, Helsinki Institute of Technology,
Finland, August 2002.
Introduction To Data Compression presented at CTVC, Syracuse, New York, January 12, 2001.
An Introduction To Genetic Algorithms
(ps.gz, 162K) lecture for the PhD students, Department of
Computer Science, University of Malaga, Spain, December 13, 2000.
Genetic Algorithms: Applications (ps, 131K) presented at the Oberseminar Theoretische Informatik, Technische Universität München, Germany, June 26, 1996.
Abstract:The presentation will start with a very
to Genetic Algorithms. It will then shift to applying these
algorithms as heurisics for tackling NP-complete
combinatorial optimization problems. More precisely, we show
how a class of these probabilistic search algorithms,
modeled after organic evolution, can be used to obtain good solutions
when applied to highly-constrained problems such as the maximum-cut,
the minimum tardy task, the set covering and the terminal assignment.
We show that even though the genetic algorithm samples but a tiny fraction of the search space, one can achieve good results by incorporating a graded penalty term in the fitness (objective) function.
Genetic Algorithms: Some Theory (ps, 119K) presented at the Oberseminar Theoretische Informatik, Technische Universität München, Germany, July 3, 1996.
Abstract:The presentation will start with a
discussion on the
Schemata Theorem. Its virtues and shortcomings will be
analyzed. We then shift to Walsh functions, which form
a convenient basis for the expansion of fitness functions.
These orthogonal, rectangular functions, have been used
to compute the average fitness values of schemata, to
decide whether a certain function is hard or easy for
a genetic algorithm, and to design deceptive functions
for the genetic algorithm.
We then investigate the use of Haar functions for the same purpose, and compare them to Walsh functions. The exploration of Fast Walsh-Haar transforms and some concluding remarks about other models for genetic algorithms will end the presentation.
Evolutionary Programming and its Applications presented at the Department of Computer Engineering, Santa Clara University, April 14, 1994.
Haar Functions in Evolution Strategies presented at the Department of Computer Science, University of Dortmund, Germany, July 16, 1993.
Genetic Algorithms and their Applications presented at the Mathematics and Computer Science Colloquium, April 22, 1993.
Evolutionary-Based Optimization Algorithms presented at the Institute of Mathematics, The Bulgarian Academy of Sciences, Sofia, Bulgaria, April 21, 1992.
Testing Algorithms presented at the Biomedical Engineering Twenty Fifth Anniversary Symposium, The Johns Hopkins School of Medicine, Baltimore, April 17-19, 1990.
Conversion Algorithms for Binary Decision Diagrams and their Efficiency invited speaker at the Logic and Computer-Aided Design Colloquim, Syracuse University, Syracuse, April 25, 1986.