Seminars

Studies on Quantile Regression

82
reads

Ming-Yueh Huang

2010-11-26
12:45:00 - 14:45:00

R440 , Astronomy and Mathematics Building

In this talk, I will give a brief discussion of estimation procedures for quantile regression models. For the case of a scalar covariate, Fan, Hu, and Truong (1994) and Yu and Jones (1997) provided several non-parametric methods to estimate conditional quantile curves. With high-dimensional covariates, Koenker and Bassett (1978) first introduced the linear regression quantiles. By extending the linear quantile model to a single-index one, Wu, Yu, and Yu (2010) developed an estimation criterion for the index-coefficients and the related inferences. Based on a more general conditional distribution model, I further propose a class of estimation procedures.