Recent Research


Much of my research deals with quantile regression. This work aims to provide a comprehensive approach to the estimation of conditional quantile functions, thus providing a more complete analysis of the stochastic relationship between economic variables than that provided by classical linear regression which restricts attention to estimation and inference about conditional mean functions. Recent work focuses more on nonparametric density estimation with penalty methods, and most recently empirical Bayes methods. References to this and earlier work may be found by perusing the following links:

Convex Optimization Survey
Empirical Bayes and Mixtures
The Ignorant Monopolist Redux
Additive Models and Confidence Bandaids
Quasi-Concave Density Estimation
Censored Quantile Regression
Bayes in the Baño
Reproducible Econometric Research
The Median is the Message
Quantile Regression Bracketology
Parametric Links for Binary Response Models
Density Estimation by Total Variation Regularization
Quantile Regression Methods for for Reference Growth Charts,
Quantile Regression Methods for Recursive Structural Equation Models,
Quantile Regression for Longitudinal Data,
Quantile Autoregression,
A Sparse Frisch-Newton Algorithm for Quantile Regression
Inequality Constrained Quantile Regression
SparseM: A Sparse Matrix Package for R
Pessimistic Portfolio Allocation and Choquet Expected Utility
Penalized Triograms: Total Variation Regularization for Bivariate Smoothing
Quantile Regression an Introduction
Inference on the Quantile Regression Process
Pathological Regression Asymptotics
Duration Models for Unemployment
Survival Analysis for Medflies
Goodness of Fit for Quantile Regression
Galton, Edgeworth, Frisch ...
Quantile Regression Software
Hare and Tortoise: l1 vs l2 Computation
Nonlinear Quantile Regression
Quantile Smoothing Splines
Weighted Quantile Regression
Confidence Intevals for Quantile Regression
Quantile Regression for ARCH Models
Rank Tests for Linear Models
Two papers on GMM
Rank Tests for Unit Roots

Send comments and suggestions to roger@ysidro.econ.uiuc.edu