scientific article; zbMATH DE number 2217287
The authors consider the model \[ y =\begin{cases} {\beta_0,} &\quad x<y, \\ {\beta_0+\beta_1(x-\gamma)+\beta_2(x-\gamma)^2}, & \quad x\geq y, \end{cases} \] where \(x\) is the indentation depth and \(y\) the applied force, for modelling so called Martens hardness, respectively, the model \[ y=\beta_0+\beta_2[x-\gamma]_+^2. \] The values of variables \(x\) and \(y\) at the measurement points are not observable. Instead, one can observe the values of two random variables \(X_i=x_i+\epsilon_{X_i}\) and \(Y_i=y_i +\epsilon_{Y_i},\;i=1,\ldots,n\), where \(\epsilon_{X_i}\) and \(\epsilon_{Y_i}\) are measurement errors. Here the parameter \(\gamma\) is of main interest while \(\beta_i\) is a nuisance parameter. The main result of the paper is the derivation of the maximum likelihood estimator and its properties suitable for the calibration process. The paper is complemented by a simulation study.
- Estimation of threshold values and regression parameters in threshold regression model
- Threshold estimation in the log-gamma model
- A smoothed least squares estimator for threshold regression models
- A threshold estimation problem for processes with hysteresis
- Semi-parametric estimation of a generalized threshold regression model under conditional quantile restriction
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 Q5699210)
