|
Activity Number:
|
130
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 3, 2009 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Business and Economic Statistics Section
|
| Abstract - #305427 |
|
Title:
|
Efficient Quantile Regression
|
|
Author(s):
|
Yoonsuh Jung*+ and Yoonkyung Lee and Steven N. MacEachern
|
|
Companies:
|
The Ohio State University and The Ohio State University and The Ohio State University
|
|
Address:
|
1958 Neil Avenue, Columbus, OH, 43210,
|
|
Keywords:
|
Quantile Estimator ; Quantile Regression
|
|
Abstract:
|
Quantile regression in a finite sample case can be made more efficient by rounding the sharp corner of the loss. The main modification involves L2 adjustment in the middle part of the loss function. The resulting modified loss has qualitatively the same shape as Huber's loss when estimating a conditional median. To achieve consistency in the large sample case, the range of L2 adjustment is controlled by a sequence which reduces to zero as the sample size increases. Without covariates, the procedure leads to a modified sample quantile estimator. In both cases, we provide conditions on the convergence rates of the sequence of modified functions for consistency. Simulation studies reveal excellent finite sample performance of L2 adjusted quantile estimators and regression quantiles under various situations.
|
