You are here:
Publication details
Kernel Regression Model with Correlated Errors
Authors | |
---|---|
Year of publication | 2014 |
Type | Chapter of a book |
MU Faculty or unit | |
Citation | |
Description | Kernel regression is one of the commonly used nonparametric methods for an estimation of a regression function. Nevertheless, there is a problem of choosing the value of the smoothing parameter, the bandwidth. In the case of independent observations the literature on the bandwidth selection is quite extensive. However, these standard methods, like cross-validation, perform badly when the errors are correlated. There are several possibilities how to overcome this. We will present and compare the partitioned cross-validation method and the plug-in method. |