********************************************************************


                                                     Seminar

             Department of Systems Engineering and Engineering Management
                                  The Chinese University of Hong Kong

------------------------------------------------------------------------------------------

　

　

　

Title

:

Efficient Derivative Free Deterministic Search Algorithms for Global Optimization of Continuous Functions

　

　

　

Speaker

:

Prof. Min Sun

　

　

Professor of Applied Mathematics, Department of Mathematics, The University of Alabama, Tuscaloosa, AL 35487, USA

　

　

currently as Visiting Professor, Department of Applied Mathematics, Hong Kong Polytechnic University

　

　

　

Date

:

May 8th, 2009 (Friday)

　

　

　

Time

:

4:00 p.m. - 5:00 p.m.

　

　

　

Venue

:

Room 401

　

　

William M.W. Mong Engineering Building

　

　

(Engineering Building Complex Phase 2)

　

　

CUHK

　

　

　

------------------------------------------------------------------------------------------

Abstract:
　

This talk starts with a general global search problem along with a prototype of general deterministic global search method, called division-and-deletion algorithm (DDA). Four types of convergence conditions for DDA are introduced. They lead to various necessary and/or sufficient conditions of convergence. Several sample realizations of the DDA prototype are then discussed for global optimization of continuous functions. They include new derivative free global optimization algorithms based on interval arithmetic that require no memory or only a limited amount of memory and estimate one or multiple globally optimal solutions. Theoretically they are shown to be able to find one or multiple globally optimal solutions under certain conditions. Numerically they have been tested with many examples of a variety of dimensions ranging from 1 to 1000 in a basic personal computer. The numerical tests have shown that the algorithms would have a good chance to uccessfully reach a good approximation of a globally optimal solution much more quickly and using much less memory space than a conventional interval method. The effects of the memory size on reliability, overall efficiency, and the number of optimal solutions are investigated.


-------------------------------------------------------------------------------------------

Biography:
　

I received my Ph.D. degree in applied mathematics from Wayne State University in Michigan in 1987. After visiting University of Houston in Texas for two years, I have been at the University of Alabama. My primary research area was in the optimal control theory and applications (deterministic, stochastic, and numerical). However since 1990 I have been developing research interests in a much wider range of areas of applied mathematics and related fields, including linear and nonlinear programming, statistics, numerical mathematics, mathematical economics, power systems, artificial intelligence,
groundwater modeling, and modeling and simulation for magnetic materials and information technology. My most recent work is on development of deterministic global optimization algorithms. I have designed fast interval-based search methods for global minimization of continuous functions. I intend to apply them to solve various optimization problems in science and engineering.


************************* ALL ARE WELCOME ************************

　

　

　

Host

:

Prof. Li Duan

Tel

:

(852) 2609-8316

Email

:

dli@se.cuhk.edu.hk

　

　

　

Enquiries

:

Prof. Nan Chen or Prof. Sean X. Zhou

　

:

Department of Systems Engineering and Engineering Management

　

　

CUHK

Website

:

http://www.se.cuhk.edu.hk/~seg5810

Email

:

seg5810@se.cuhk.edu.hk

　

　

　

********************************************************************