Estimation in autoregressive model with measurement error

Abstract: Consider an autoregressive model with measurement error: we observe

Z_{i} = X_{i} + e p s i l o n_{i}

, where

X_{i}

is a stationary solution of the equation

X_{i} = f_{h e t a^{0}} (X_{i - 1}) + x i_{i}

. The regression function

f_{h e t a^{0}}

is known up to a finite dimensional parameter

h e t a^{0}

. The distributions of

X_{0}

and

x i_{1}

are unknown whereas the distribution of

e p s i l o n_{1}

is completely known. We want to estimate the parameter

h e t a^{0}

by using the observations

Z_{0}, .., Z_{n}

. We propose an estimation procedure based on a modified least square criterion involving a weight function

w

, to be suitably chosen. We give upper bounds for the risk of the estimator, which depend on the smoothness of the errors density

f_{e} p s i l o n

and on the smoothness properties of

w f_{h} e t a

.

Recommendations

Cited In (15)

This page was built for publication: Estimation in autoregressive model with measurement error