The least trimmed quantile regression
نویسندگان
چکیده
The linear quantile regression estimator is very popular and widely used. It is also well known that this estimator can be very sensitive to outliers in the explanatory variables. In order to overcome this disadvantage, the usage of the least trimmed quantile regression estimator is proposed to estimate the unknown parameters in a robust way. As a prominent measure of robustness, the breakdown point of this estimator is characterized and its consistency is proved. The performance of this approach in comparison with the classical one is illustrated by an example and simulation studies.
منابع مشابه
The Symmetric Type Two-stage Trimmed Least Squares Estimator for the Simultaneous Equations Model
A two-stage symmetric regression quantile is considered as an alternative for estimating the population quantile for the simultaneous equations model. We introduce a two-stage symmetric trimmed least squares estimator (LSE) based on this quantile. It is shown that, under mixed multivariate normal errors, this trimmed LSE has asymptotic variance much closer to the Cramér-Rao lower bound than som...
متن کاملRegression quantiles and trimmed least squares estimator under a general design
where X = (X1, ...,X„)' is the vector of independent observations, C is the n x p design matrix, /} = (ftu ...,PP)' is the vector of unknown parameters and £ = = ( £ ] , . . . , E„)' where Eu ..., E„ are independent and identically distributed (i.i.d.) random variables with a continuous distribution function (d.f.) F. Our main interest is in robust estimating the parameter /?. For the location ...
متن کاملEditorial for the special issue on quantile regression and semiparametric methods
Quantile regression and other semiparametric models have been widely recognized as important data analysis tools in statistics and econometrics. Thesemethods donot rely strictly onparametric likelihoodbut avoid the curse of dimensionality associated with many nonparametric models. The journal Computational Statistics and Data Analysis regularly publishes papers on these semiparametric methods, ...
متن کاملEmpirical Regression Quantile
This study proposes a new use of goal programming for empirically estimating a regression quantile hyperplane. The approach can yield regression quantile estimates that are less sensitive to not only non-Gaussian error distribut.ions but also a small sample size t.han conventional regression quantile methods. The performance of regression quantile estimates is compared with least absolute value...
متن کاملBagging Binary and Quantile Predictors for Time Series: Further Issues
Bagging (bootstrap aggregating) is a smoothing method to improve predictive ability under the presence of parameter estimation uncertainty and model uncertainty. In Lee and Yang (2006), we examined how (equal-weighted and BMA-weighted) bagging works for onestep ahead binary prediction with an asymmetric cost function for time series, where we considered simple cases with particular choices of a...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
- Computational Statistics & Data Analysis
دوره 56 شماره
صفحات -
تاریخ انتشار 2012