Monte Carlo algorithms. (Q616584): Difference between revisions

This monograph, which is based on lectures held by the authors at several German universities, gives a comprehensive introduction to Monte Carlo methods. The first chapters present the basic ideas and concepts of Monte Carlo methods for direct simulation. After theoretical results like convergence proofs, limit theorems and error estimates, each chapter gives practical examples from particle physics and financial and actuarial mathematics and finally ends with a collection of exercises left to the reader. Chapter 4 discusses the simulation of non-uniform distributions (via the inversion and the acceptance-rejection methods) and the simulation of various stochatic processes (time-discrete Markov processes, Brownian motion, Poisson process, etc.). Chapter 5 presents the usual techniques for variance reduction to improve convergence rates. The two final chapters of the monograph discuss the Markov chain Monte Carlo method and the general concepts of numerical integration. Starting with deterministic algorithms and their shortcomings, randomized algorithms for numerical integration are investigated thoroughly and their advantages over deterministic methods are clearly shown. These two chapters also give references to current research for further reading. All in all, this monograph by three well-known researchers in the field gives a well-founded mathematical introduction to Monte Carlo methods, suited for students as well as for researchers and even practitioners with some mathematical background.