Loaded dice in Monte Carlo : importance sampling in phase space integration and probability distributions for discrepancies

Discrepancies play an important role in the study of uniformity properties of point sets. Their probability distributions are a help in the analysis of the efficiency of the Quasi Monte Carlo method of numerical integration, which uses point sets that are distributed more uniformly than sets of independently uniformly distributed random points. In this thesis, generating functions of probability distributions of quadratic discrepancies are calculated using techniques borrowed from quantum field theory. The second part of this manuscript deals with the application of the Monte Carlo method to phase space integration, and in particular with an explicit example of importance sampling. It concerns the integration of differential cross sections of multi-parton QCD-processes, which contain the so-called kinematical antenna pole structures. The algorithm is presented and compared with RAMBO, showing a substantial reduction in computing time. In behalf of completeness of the thesis, short introductions to probability theory, Feynman diagrams and the Monte Carlo method of numerical integration are include