17-28 June 2002
Organizer: Pierre L'Écuyer
The aim of this workshop is to bring together the world-leading experts on the theoretical and practical aspects of random number generation by computer and the design of highly uniform point sets for quasi-Monte Carlo integration. The workshop is spread over a period of two weeks, to allow enough time for serious discussions and collaborative work. We look forward to a fruitful exchange between participants from different backgrounds, such as mathematicians, physicists, computer scientists, statisticians, and so on. The general theme is the development of practical random number generation software for various classes of applications, such as simulation, statistics, numerical analysis, computer games, lotteries, cryptology, etc. Those application often have different requirements, which in a sense makes things more interesting. We are concerned with mathematical tools for selecting and testing specific generators, the theoretical and statistical analysis of families of generators, software implementation techniques, and all related work that could be useful for the construction or analysis of practical generators. Uniform and non-uniform random number generation are both covered. Highly uniform (or quasi-Monte Carlo) (QMC) point sets and sequences are designed to be more uniform than typical random points. In several simulation setups, they can advantageously replace the numbers produced by traditional random number generators (RNGs). Their construction and analysis can be based on very similar ideas and tools than for RNGs. One of our goals is to strenghten this connection. Topics of interest: We plan to limit the talks to at most five hours a day, to allow time for collaborative work. Each participant will receive a booklet of abstracts. A special issue of the ACM Transactions on Modeling and Computer Simulation will be devoted to the best papers coming out of this workshop. The submission deadline will be announced at the workshop.
|