Title :
Simulation from non-standard distributions using envelope methods
Author :
Evans, Michael J. ; Swartz, Tim B.
Author_Institution :
Dept. of Stat., Toronto Univ., Ont., Canada
Abstract :
This paper considers the development of envelope methods as a tool for simulation. Envelope methods are based on the construction of simple envelopes to functions. The proposed envelopes are general, require little input from the user and are based on the concavity structure of the function or some transformation of the function. The construction of these envelopes facilitates variate generation using the adaptive rejection algorithm
Keywords :
simulation; stochastic processes; adaptive rejection algorithm; concavity structure; envelope methods; non-standard distributions simulation; variate generation; Ear; Higher order statistics; Integral equations; Mathematics; Packaging; Statistical distributions; Stochastic systems; Tail; Upper bound;
Conference_Titel :
Simulation Conference, 2000. Proceedings. Winter
Conference_Location :
Orlando, FL
Print_ISBN :
0-7803-6579-8
DOI :
10.1109/WSC.2000.899765