Abstract:A sharp, distribution free, non-asymptotic result is proved for the concentration of a random function around the mean function, when the randomization is generated by a finite sequence of independent data and the random functions satisfy uniform bounded variation assumptions. The specific motivation for the work comes from the need for inference on the distributional impacts of social policy intervention. However, the family of randomized functions that we study is broad enough to cover wide-ranging applications. For example, we provide a Kolmogorov-Smirnov like test for randomized functions that are almost surely Lipschitz continuous, and novel tools for inference with heterogeneous treatment effects. A Dvoretzky-Kiefer-Wolfowitz like inequality is also provided for the sum of almost surely monotone random functions, extending the famous non-asymptotic work of Massart for empirical cumulative distribution functions generated by i.i.d. data, to settings without micro-clusters proposed by Canay, Santos, and Shaikh. We illustrate the relevance of our theoretical results for applied work via empirical applications. Notably, the proof of our main concentration result relies on a novel stochastic rendition of the fundamental result of Debreu, generally dubbed the "gap lemma," that transforms discontinuous utility representations of preorders into continuous utility representations, and on an envelope theorem of an infinite dimensional optimisation problem that we carefully construct.

A multivariate central limit theorem for randomized orthogonal array sampling designs in computer experiments

A Uniform Bound on a Combinatorial Central Limit Theorem for Randomized Orthogonal Array Sampling Designs

A GENERAL THEORY FOR ORTHOGONAL ARRAY BASED LATIN HYPERCUBE SAMPLING

Central Limit Theorems for Four New Types of U-Designs

The Asymptotic Distribution and Berry–Esseen Bound of a New Test for Independence in High Dimension with an Application to Stochastic Optimization

Orthogonal Array-Based Latin Hypercubes

Approximation of Bounds on Mixed Level Orthogonal Arrays

An Improvement of Normal Approximation of Randomized Orthogonal Array Sampling Designs

Some New Classes of Orthogonal Latin Hypercube Designs

Sliced Orthogonal Array-Based Latin Hypercube Designs

Explicit Orthogonal Arrays and Universal Hashing with Arbitrary Parameters

Orthogonal Array Based Locally D-Optimal Designs for Binary Responses in the Presence of Factorial Effects

How One-Factor-at-a-Time Experimentation Can Lead to Greater Improvements Than Orthogonal Arrays

Construction of orthogonal maximin distance designs

Adaptive and Oblivious Randomized Subspace Methods for High-Dimensional Optimization: Sharp Analysis and Lower Bounds

A New Class of Latin Hypercube Designs with High-Dimensional Hidden Projective Uniformity

Nested Lattice Sampling: A New Sampling Scheme Derived by Randomizing Nested Orthogonal Arrays

Complete enumeration of two-Level orthogonal arrays of strength $d$ with $d+2$ constraints

Concentration of Randomized Functions of Uniformly Bounded Variation

Construction of Sliced Space-Filling Designs Based on Balanced Sliced Orthogonal Arrays

Construction of orthogonal-MaxPro Latin hypercube designs