Modeling dependence in simulation input: The case for copulas

We discuss copulas for incorporating dependence in the input distributions to a simulation model. We start by motivating the need for incorporating dependence in the primitive inputs to a simulation. Copulas are then introduced as a convenient and flexible model to incorporate dependence. We rigorou...

Full description

Saved in:
Bibliographic Details
Main Authors: Pasupathy, Raghu, Nagaraj, Kalyani
Format: Conference Proceeding
Language:English
Subjects:
Computational modeling
Computer simulation
Conferences
Data models
Distribution functions
Exposure
Inference
Mathematical model
Mathematical models
Modelling
Portfolios
Production
Tutorials
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We discuss copulas for incorporating dependence in the input distributions to a simulation model. We start by motivating the need for incorporating dependence in the primitive inputs to a simulation. Copulas are then introduced as a convenient and flexible model to incorporate dependence. We rigorously define copulas, introduce some of their basic properties, illustrate popular copula families, and discuss methods for copula estimation. Since this is an introductory tutorial, we have attempted to keep all exposition at a basic mathematical level without omitting important technical details. The oral presentation of this tutorial will include additional discussion on random variate generation, copula inference, and tail dependence.
ISSN:1558-4305
DOI:10.1109/WSC.2015.7408300