Automatic Nonuniform Random Variate Generation

Wolfgang Hörmann, Josef Leydold, Gerhard Derflinger - Collection Statistics and Computing

442 pages, parution le 05/11/2003

Ajouter à une liste

Indisponible

Résumé

Non-uniform random variate generation is an established research area in the intersection of mathematics, statistics and computer science. Although random variate generation with popular standard distributions have become part of every course on discrete event simulation and on Monte Carlo methods, the recent concept of universal (also called automatic or black-box) random variate generation can only be found dispersed in literature. This new concept has great practical advantages that are little known to most simulation practitioners. Being unique in its overall organization the book covers not only the mathematical and statistical theory but also deals with the implementation of such methods.
All algorithms introduced in the book are designed for practical use in simulation and have been coded and made available by the authors. Examples of possible applications of the presented algorithms (including option pricing, VaR and Bayesian statistics) are presented at the end of the book.

Written for:
Researchers and graduate students in Computational Statistics

Part I Preliminaries
- Introduction
- General Principles in Random Variate Generation
- General Principles for Discrete Distributions
Part II Continuous Univariate Distributions
- Transformed Density Rejection (TDR)
- Strip Methods
- Methods Based on General Inequalities
- Numerical Inversion
- Comparison and General Considerations
- Distributions Where the Density Is Not Known Explicitly
Part III Discrete Univariate Distributions
- Discrete Distributions
Part IV Random Vectors
- Multivariate Distributions
Part V Implicit Modeling
- Combination of Generation and Modeling
- Time Series (Authors Michael Hauser and Wolfgang Hörmann)
- Markov Chain Monte Carlo Methods
- Some Simulation Examples
- List of Algorithms
- References

Caractéristiques techniques

	PAPIER
Éditeur(s)	Springer
Auteur(s)	Wolfgang Hörmann, Josef Leydold, Gerhard Derflinger
Collection	Statistics and Computing
Parution	05/11/2003
Nb. de pages	442
Format	16 x 24
Couverture	Relié
Poids	780g
Intérieur	Noir et Blanc
EAN13	9783540406525