NONLINEAR STATE-SPACE MODELS WITH PARTIALLY SPECIFIED DISTRIBUTIONS ON STATES

被引：15

作者：

SMITH, JQ

机构：

来源：

JOURNAL OF FORECASTING | 1990年 / 9卷 / 02期

关键词：

Bayesian forecasting; Conditional independence; Dynamic linear model; Identifiability; Invariant properties; Preservation of constraints; Steady model;

D O I：

10.1002/for.3980090206

中图分类号：

F [经济];

学科分类号：

02 ;

摘要：

The author argues, through the discussion of several examples, why models like the power steady model (Smith 1979, 1981a) are attractive not only because of the simplicity of their one‐step‐ahead forecast distributions but also because they can be justified through sets of properties that any gradually evolving process may be expected to satisfy. Copyright © 1990 John Wiley & Sons, Ltd.

引用

页码：137 / 149

页数：13

共 19 条

[1] MARKOVIAN REPRESENTATION OF STOCHASTIC-PROCESSES AND ITS APPLICATION TO ANALYSIS OF AUTOREGRESSIVE MOVING AVERAGE PROCESSES [J].

AKAIKE, H .

ANNALS OF THE INSTITUTE OF STATISTICAL MATHEMATICS, 1974, 26 (03) :363-387

[2]

AMEEN J, 1985, BAYESIAN STATISTICS, V2

[3]

AMEEN JRM, 1985, 78 U WARW DEP STAT R

[4]

ATTWELL DN, 1988, 137 U WARW DEP STAT

[5]

ATTWELL DN, 1988, 145 U WARW DEP STAT

[6]

HARRISON PJ, 1976, J R STAT SOC B, V38, P205

[7]

MIGON HS, 1985, BAYESIAN STATISTICS, V2

[8]

Moran P. A. P., 1968, INTRO PROBABILITY TH

[9]

Pole A., 1988, 136 U WARW DEP STAT, P136

[10]

SMITH JQ, 1981, J R STAT SOC B, V43, P256

← 1 2 →