MULTINOMIAL LOGISTIC-REGRESSION ALGORITHM

被引：376

作者：

BOHNING, D ^{[1
]}

机构：

[1] FREE UNIV BERLIN,DEPT EPIDEMIOL,W-1000 BERLIN 45,GERMANY

来源：

ANNALS OF THE INSTITUTE OF STATISTICAL MATHEMATICS | 1992年 / 44卷 / 01期

关键词：

KRONECKER PRODUCT; LOEWNER ORDERING; LOWER BOUND PRINCIPLE; MONOTONICITY;

D O I：

10.1007/bf00048682

中图分类号：

O21 [概率论与数理统计]; C8 [统计学];

学科分类号：

020208 ; 070103 ; 0714 ;

摘要：

The lower bound principle (introduced in Bohning and Lindsay (1988, Ann. Inst. Statist. Math., 40, 641-663), Bohning (1989, Biometrika, 76, 375-383) consists of replacing the second derivative matrix by a global lower bound in the Loewner ordering. This bound is used in the Newton-Raphson iteration instead of the Hessian matrix leading to a monotonically converging sequence of iterates. Here, we apply this principle to the multinomial logistic regression model, where it becomes specifically attractive.

引用

页码：197 / 200

页数：4

共 3 条

[1] A NOTE ON THE MATRIX ORDERING OF SPECIAL C-MATRICES
BAKSALARY, JK
PUKELSHEIM, F
[J]. LINEAR ALGEBRA AND ITS APPLICATIONS, 1985, 70 (OCT) : 263 - 267
[2] LIKELIHOOD INFERENCE FOR MIXTURES - GEOMETRICAL AND OTHER CONSTRUCTIONS OF MONOTONE STEP-LENGTH ALGORITHMS
BOHNING, D
[J]. BIOMETRIKA, 1989, 76 (02) : 375 - 383
[3] MONOTONICITY OF QUADRATIC-APPROXIMATION ALGORITHMS
BOHNING, D
LINDSAY, BG
[J]. ANNALS OF THE INSTITUTE OF STATISTICAL MATHEMATICS, 1988, 40 (04) : 641 - 663

← 1 →