MCQMC 2020, Minisymposium Random points: Quality Criteria and Applications
Sprache des Titels:
Englisch
Original Kurzfassung:
Since the invention of the Monte Carlo method by Metropolis, Ulam, and von Neumann, random and pseudo-random point sets play an important role in stochastic simulation, numerical integration, optimization and
other areas of applied mathematics. In all these areas and in applications in the natural sciences and in computer science there are usually different requirements that ?good point sets? should satisfy. Quality criteria of common interest comprise, e.g., small variance (in stochastic simulation), low discrepancy (in quasi-Monte
Carlo (QMC) integration) or small dispersion (in global optimization as, e.g., hyperparameter optimization in Deep Learning). Rather new criteria in stochastic simulation and discrepancy theory are based on notions of
negative dependent random variables. In this special session we want to discuss different quality criteria for random or pseudo-random point sets, the construction of good point sets, and their performance in applications.