On a model with errors in variables described by stochastic differential equations (Q3707192)
From MaRDI portal
!
This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:
scientific article; zbMATH DE number 3936297
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | On a model with errors in variables described by stochastic differential equations |
scientific article; zbMATH DE number 3936297 |
Statements
On a model with errors in variables described by stochastic differential equations (English)
0 references
1985
0 references
errors-in-variables models
0 references
consistency
0 references
dynamic system
0 references
Ornstein- Uhlenbeck process
0 references
minimum-contrast-estimator
0 references
0.7597396969795227
0 references
0.7545890808105469
0 references
0.754431962966919
0 references
0.7532321214675903
0 references
We consider a dynamic system, described by the stochastic differential equation \[ d\xi_ t=\vartheta \xi_ tdt+dW_{1t},\quad \xi_ 0=0,\quad \vartheta <0. \] Because of the noise-error, we observe not \((\xi_ t)_{t\geq 0}\), but \((y_ t)_{t\geq 0}\) connected with \((\xi_ t)_{t\geq 0}\) as follows: \(y_ t=\xi_ t+\epsilon_ t\), \(t\geq 0\). The process \((\epsilon_ t)_{t\geq 0}\) is an Ornstein- Uhlenbeck process: \[ d\epsilon_ t=-\alpha \epsilon_ tdt+dW_{2t},\quad \alpha >0,\quad \epsilon_ 0\sim N(0,1/2\alpha) \] and \((W_{1t})_{t\geq 0}\) and \((W_{2t})_{t\geq 0}\) are independent. We show, that in this case the estimator \[ {\tilde \vartheta}_ T=(\int^{T}_{0}y^ 2_ tdt)^{-1}\cdot \int^{T}_{0}y_ tdy_ t \] is not consistent and propose a method for constructing an alternative minimum-contrast-estimator of \(\vartheta\).
0 references