Optimal Bandwidth Selection for Nonparametric Conditional Distribution and Quantile Functions

We propose a data-driven least squares cross-validation method to optimally select smoothing parameters for the nonparametric estimation of conditional cumulative distribution functions and conditional quantile functions. We allow for general multivariate covariates that can be continuous, categorical or a mix of either. We provide asymptotic analysis, examine finite-sample properties via Monte Carlo simulation, and consider an application involving testing for first order stochastic dominance of children's health conditional on parental education and income.