Stratified incomplete local simplex tests for curvature of nonparametric multiple regression
arXiv:2003.09091 · doi:10.3150/22-BEJ1459
Abstract
Principled nonparametric tests for regression curvature in are often statistically and computationally challenging. This paper introduces the stratified incomplete local simplex (SILS) tests for joint concavity of nonparametric multiple regression. The SILS tests with suitable bootstrap calibration are shown to achieve simultaneous guarantees on dimension-free computational complexity, polynomial decay of the uniform error-in-size, and power consistency for general (global and local) alternatives. To establish these results, a general theory for incomplete -processes with stratified random sparse weights is developed. Novel technical ingredients include maximal inequalities for the supremum of multiple incomplete -processes.
arXiv commentAccepted to Bernoulli
References in corpus (12)
- Uniform in bandwidth consistency of kernel-type function estimators
- Gaussian approximation of suprema of empirical processes
- A new perspective on least squares under convex constraint
- On local -statistic processes and the estimation of densities of functions of several sample variables
- Approximating high-dimensional infinite-order -statistics: statistical and computational guarantees
- Multivariate convex regression: global risk bounds and adaptation
- Optimality of Maximum Likelihood for Log-Concave Density Estimation and Bounded Convex Regression
- A Projection Framework for Testing Shape Restrictions That Form Convex Cones
- Nonparametric Least Squares Estimation of a Multivariate Convex Regression Function
- On Convex Least Squares Estimation when the Truth is Linear
- Testing nonparametric shape restrictions
- Stratified incomplete local simplex tests for curvature of nonparametric multiple regression