In a wide variety of engineering applications based on probability distribution models, the adequate choice of these models is critical. This paper introduces a piecewise multilinear (PML) function model for probability densities. The model complexity can be tuned to the application at hand. Deriving the PML model from a given pdf as well as generating random samples satisfying this pdf, appear to be easy.
Nyssen, E & Jacquet, W 2008, Generating a multivariate random sample, using a pdf model based on a piecewise multilinear approximation. in Abstract book of the 22nd national conference of the Belgian Operations Research Society - ORBEL 22. Abstract book of the 22nd national conference of the Belgian Operations Research Society - ORBEL 22, pp. 32-33.
Nyssen, E., & Jacquet, W. (2008). Generating a multivariate random sample, using a pdf model based on a piecewise multilinear approximation. In Abstract book of the 22nd national conference of the Belgian Operations Research Society - ORBEL 22 (pp. 32-33). (Abstract book of the 22nd national conference of the Belgian Operations Research Society - ORBEL 22).
@inbook{4918f1872bac4cbf85946fcf5bd5a7f5,
title = "Generating a multivariate random sample, using a pdf model based on a piecewise multilinear approximation",
abstract = "In a wide variety of engineering applications based on probability distribution models, the adequate choice of these models is critical. This paper introduces a piecewise multilinear (PML) function model for probability densities. The model complexity can be tuned to the application at hand. Deriving the PML model from a given pdf as well as generating random samples satisfying this pdf, appear to be easy.",
keywords = "probability density function, random number generation, stochastic simulation",
author = "Edgard Nyssen and Wolfgang Jacquet",
year = "2008",
month = jan,
language = "English",
series = "Abstract book of the 22nd national conference of the Belgian Operations Research Society - ORBEL 22",
pages = "32--33",
booktitle = "Abstract book of the 22nd national conference of the Belgian Operations Research Society - ORBEL 22",
}