Advances and Applications in Statistics
Volume 52, Issue 3, Pages 203 - 213
(March 2018) http://dx.doi.org/10.17654/AS052030203 |
|
ON COMPARISON OF ESTIMATION METHODS IN QUANTILE REGRESSION
Song Jea Woo and Kee-Hoon Kang
|
Abstract: As in the case of economics or clinical trial data, it is often the case that the distribution of extreme values, such as the tail, is of more concern than the expected value. In such cases, it is common to consider quantile regression analysis rather than traditional regression analysis. In this study, we summarize the estimation methods of parametric and nonparametric approaches that can be considered in quantile regression analysis. We compare the performance of each method by using several simulations. |
Keywords and phrases: bandwidth, double kernel estimation, local linear estimation, nonparametric estimation, parametric estimation, plug-in. |
|
Number of Downloads: 450 | Number of Views: 3363 |
|