scientific article; zbMATH DE number 1046593
Let \(f(z)\) be a positive function. The following regression is under study: \[ y_s=f(x_s)+\sigma_s\xi_s, \quad s=1,\dots ,N, \] where \(\xi_s\) are noncorrelated random variables with \({\mathbf E} \xi_s=0\) and \({\mathbf D}\xi_s=1\). The problem is to estimate the function \(f\), given \(\{x_s\}\) and \(\{y_s\}\). In the article, the kernel estimator \[ \widehat f_N(z,h)=\sum_{s=1}^N\frac{\delta(h,z,x_s)}{B(N,h)}y_s \] is considered, where \(B(N,h)=\sum_{s=1}^N\delta(h,z,x_s)\) is a normalizing coefficient and \(\delta(h,z,x)\) is a kernel. In Theorems 2 and 3, necessary and sufficient conditions are given under which \(\widehat f_N(z,h)\to f(z)\) in \(L_2\) as \(h\to 0\). In Theorem 5, asymptotic normality of the estimate \(\widehat f_N(z,h)\) is studied.
- Nonparametric identification and estimation of sample selection models under symmetry
- Identification problem of the exponential tilting estimator under misspecification
- Asymptotics for estimation and testing procedures under loss of identifiability
- Asymptotic Properties for a Class of Partially Identified Models
- scientific article; zbMATH DE number 1046594 (Why is no real title available?)
- Analyzing iterations in identification with application to nonparametric \(\mathcal H_{\infty}\)-norm estimation
This page was built for publication:
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4348909)
