Distributional Random Forests for Complex Survey Designs on Reproducing Kernel Hilbert Spaces

TL;DR

This paper introduces a survey-calibrated distributional random forest method for estimating conditional distributions under complex survey designs, providing theoretical guarantees and practical applications in health data analysis.

Contribution

It develops a novel distributional random forest framework that accounts for survey design features, with proven consistency and applicability to real-world complex survey data.

Findings

Finite sample performance demonstrated through simulations.

First model-free estimation of conditional distributions under survey designs.

Application to NHANES data for subgroup risk profiling.

Abstract

We study estimation of the conditional law $P(Y|X=x)$ and continuous functionals $\Psi(P(Y|X=x))$ when $Y$ takes values in a locally compact Polish space, $X \in \mathbb{R}^p$, and the observations arise from a complex survey design. We propose a survey-calibrated distributional random forest (SDRF) that incorporates complex-design features via a pseudo-population bootstrap, PSU-level honesty, and a Maximum Mean Discrepancy (MMD) split criterion computed from kernel mean embeddings of H\'{a}jek-type (design-weighted) node distributions. We provide a framework for analyzing forest-style estimators under survey designs; establish design consistency for the finite-population target and model consistency for the super-population target under explicit conditions on the design, kernel, resampling multipliers, and tree partitions. As far as we are aware, these are the first results on model-free estimation of conditional distributions under survey designs. Simulations under a stratified two-stage cluster design provide finite sample performance and demonstrate the statistical error price of ignoring the survey design. The broad applicability of SDRF is demonstrated using NHANES: We estimate the tolerance regions of the conditional joint distribution of two diabetes biomarkers, illustrating how distributional heterogeneity can support subgroup-specific risk profiling for diabetes mellitus in the U.S. population.