An approximation of multivariate normal orthant probabilities of dimension 4: A contingency-table approach

TY - JOUR

T1 - An approximation of multivariate normal orthant probabilities of dimension 4

T2 - A contingency-table approach

AU - Wang, Yuchung J.

PY - 1987/1/1

Y1 - 1987/1/1

N2 - The collection of all orthant probabilities for a four-dimensional multivariate normal distribution can be considered as a 2x2x2x2 contingency table. There are no three-way interactions for the contingency table. And the fourth-order interaction is negligible. The six two-way marginal probabilities are sufficient statistics for approximating the 24 orthant probabilities. The iterative proportional fitting algorithm is used to compute the four-dimensional contingency table from the six marginal 2x2 tables.

AB - The collection of all orthant probabilities for a four-dimensional multivariate normal distribution can be considered as a 2x2x2x2 contingency table. There are no three-way interactions for the contingency table. And the fourth-order interaction is negligible. The six two-way marginal probabilities are sufficient statistics for approximating the 24 orthant probabilities. The iterative proportional fitting algorithm is used to compute the four-dimensional contingency table from the six marginal 2x2 tables.

KW - cross-product ratio

KW - interactions

KW - iterative proportional fitting algorithm

KW - log-linear model

UR - https://www.scopus.com/pages/publications/0040125197

UR - https://www.scopus.com/pages/publications/0040125197#tab=citedBy

U2 - 10.1080/03610918708812635

DO - 10.1080/03610918708812635

M3 - Article

SN - 0361-0918

VL - 16

SP - 1039

EP - 1049

JO - Communications in Statistics - Simulation and Computation

JF - Communications in Statistics - Simulation and Computation

IS - 4

ER -