Stochastic Control, Communications and Numerics Perspective
Adelaide, September 29, 30 and October 1, 2004
Sponsored by the University of South Australia and AMSI
Postgraduate Student Scholarship
Postgraduate Student Scholarship form
Structure of the Scientific Program
Jerzy A. Filar
Foundation Chair of Mathematics and Statistics
Centre for Industrial and Applied Mathematics
University of South Australia
Email: J.Filar@unisa.edu.au
Vladimir G. Gaitsgory
Associate Professor of Mathematics
Centre for Industrial and Applied Mathematics
University of South Australia
Email: V.Gaitsgory@unisa.edu.au
Fima Klebaner
Professor of Statistics
School of Mathematical Sciences
Monash University, Clayton, Vic. 3800
In recent years many problems arising naturally in control, signal processing, communications and operations research required the understanding of uncertainty at a sophisticated level. This need stimulated advances in both theoretical stochastic methods and in their novel applications. While these advances rest upon solid foundations of the theories of probability, stochastic processes and diffusions, their manifold realizations often take novel forms stemming from the underlying motivating context that is frequently that of an engineering problem. Hence the body of knowledge that has been emerging is a characteristic blend of theory, approximation methods, numerical algorithms and applications. It differentiates itself from some of the more classical theory of stochastic processes precisely by this hybrid nature: all four aspects seem indispensable and constantly reinforce one another.
One of the difficulties affecting this field is its “cultural fragmentation”. Mathematicians, statisticians, control and communications engineers and operations researchers often analyze similar problems using conceptually similar techniques but employ sufficiently different terminology and notation to make each other’s contributions difficult to access and appreciate. This workshop is designed to help interested researchers and students overcome these barriers and to focus their attention on some of the most exciting current lines of investigation.
Postgraduate Student Scholarships
A limited number of $500 scholarships are available to postgraduate students from AMSI affiliated member institutions. Please send a short CV to Professor Jerzy Filar (jerzy.filar@unisa.edu.au) by 5pm Friday September 10th.
Postgraduate Student Posters
Postgraduate students planning to participate are invited to submit poster presentations that will be displayed at the workshop venue. Posters should be submitted by 5pm Friday September 24th to Deborah Downes (deborah.downes@unisa.edu.au )
Please click here for further details and the Postgraduate student Scholarship information sheet
Structure of the scientific program
The focus of the workshop is on mathematical methods. The orientation
is towards applications in control, communications and operations
research. The workshop will include participation by some 18 academics
and 16 postgraduate students, half of whom will be coming from
interstate.
The workshop will contain two types of lectures given by leading
international and national researchers in the field:
a) Tutorial or expository lectures intended to give the audience a serious overview of an important, still current, line of research.
and
b) Specialist lectures reporting on recent advances that the presenter considers exciting.
The two tutorial lectures will each be of 3 hours duration (including breaks) and will cover the following subtopics:
- Markov chain approximations for optimal control: construction, computation, and convergence.
P. Dupuis, Brown University.- Matrix Analytic Methods in Stochastic Modelling. N. Bean, University of Adelaide & P. Taylor, University of Melbourne
Specialist lectures reporting on recent advances and open problems that the presenters consider exciting. Typically, these will be 50 minutes in length. The provisional list of specialist presentations is given below.
- Directed Graphs, Hamiltonicity and Doubly Stochastic Matrices.
V.S. Borkar, Tata Institute, Mumbai.- Stochastic Control for Sensor Scheduling and Low Data Rate Systems.
R. Evans, University of Melbourne.- Applications of Constrained Semi-Markov Decision Processes to Deterministic Scheduling: Generalized Pinwheel Problem.
E. Feinberg, SUNY at Stony Brook.- Stochastic Embedding of Certain Discrete Optimisation Problems.
J. Filar, University of South Australia.- Limit Occupational Measures Set and Singularly Perturbed Control Systems.
V. Gaitsgory, University of South Australia.- Regularity of Markov Jump Processes
Kais Hamza, Monash University.- Priority in queues.
M. Haviv, Hebrew University of Jerusalem.- An Application of a Matrix Reduction Method in Stochastic Modelling.
P.G. Howlett, University of South Australia.- TBA.
Matthew James, ANU.- Karhunen-Loeve Approximations of Wiener Processes and Solutions of SDE’s.
S.K. Lucas, University of South Australia.- Finite Horizon Robust State Estimation for Uncertain Finite-Alphabet Hidden Markov Models with Regular Conditional Relative Entropy Constraints.
I.R. Petersen, UNSW / ADFA.- Costs and Decisions in Populations Models.
Phil Pollett & Joshua Ross, University of Queensland.International Theme Leader
We are delighted that Professor Paul Dupuis from Brown University, Rhode Island, has agreed to act as our international theme leader for the theme "Numerical Methods of Stochastic Control”. Paul Dupuis is Professor and Associate Chair of the Division of Applied Mathematics at Brown University. Prior to joining the Brown faculty in 1991, he served as Assistant Professor at the University of Massachusetts and as Visiting Assistant Professor at the Lefschetz Center for Dynamical Systems at Brown. Professor Dupuis’ research interests are in probability, computational methods, control, and related PDEs. He has written over 50 journal articles and has co-authored two books, "Numerical Methods for Stochastic Control Problems in Continuous Time" and "A Weak Convergence Approach to the Theory of Large Deviations."