Bayesian Non-parametric Simultaneous Quantile Regression for Complete and Grid Data

TL;DR

This paper introduces Bayesian non-parametric methods for simultaneous quantile regression using B-spline basis functions, capable of estimating entire quantile functions for complete and grid data, ensuring monotonicity and applicability to real-world datasets.

Contribution

It proposes novel Bayesian non-parametric approaches that estimate whole quantile functions with monotonicity constraints, applicable to both complete and grid data, improving over existing methods.

Findings

Effective in estimating quantile functions for income and hurricane data.

Maintains monotonicity of quantile levels in estimation.

Performs well in simulation studies for both data types.

Abstract

In this paper, we consider Bayesian methods for non-parametric quantile regressions with multiple continuous predictors ranging values in the unit interval. In the first method, the quantile function is assumed to be smooth over the explanatory variable and is expanded in tensor product of B-spline basis functions. While in the second method, the distribution function is assumed to be smooth over the explanatory variable and is expanded in tensor product of B-spline basis functions. Unlike other existing methods of non-parametric quantile regressions, the proposed methods estimate the whole quantile function instead of estimating on a grid of quantiles. Priors on the B-spline coefficients are put in such a way that the monotonicity of the estimated quantile levels are maintained unlike local polynomial quantile regression methods. The proposed methods have also been modified for quantile grid data where only the percentile range of each response observations are known. Simulations studies have been provided for both complete and quantile grid data. The proposed method has been used to estimate the quantiles of US household income data and North Atlantic hurricane intensity data.