Pith. sign in

REVIEW 3 major objections 6 minor 106 references

Distributional Random Forests for Complex Survey Designs

T0 review · 3 major / 6 minor · reviewed 2026-08-03 · deepseek-v4-flash

Pith's one-line read This paper introduces a survey-calibrated distributional random forest and proves it consistently estimates the full conditional distribution of a multivariate outcome under complex survey designs, with plug-in consistency for functionals s

desk verdict Useful algorithm, but the main consistency theorem is built on an unproved assumption and a mis-specified target for continuous covariates; worth engaging, needs a major revision. read the letter →

arxiv 2512.08179 v3 pith:LAESRMOO submitted 2025-12-09 stat.ME stat.ML

classification stat.MEstat.ML MSC 62D0562G0862G20
keywords distributionalrandomforestcomplexsurveydesignskernelmeanembeddingmaximumdiscrepancydesignconsistencymodelpseudo-populationbootstrapconditionaldistribution
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Complex survey designs introduce unequal selection, stratification, and clustering, so standard machine-learning distribution estimators that treat rows as i.i.d. target the wrong population. This paper introduces a survey-calibrated distributional random forest (SDRF) whose split rule compares Hájek-weighted kernel mean embeddings of child-node distributions and whose bagging step mimics the sampling design through a pseudo-population bootstrap. The central claim is that, under explicit design, kernel, resampling, and tree-partition conditions, the SDRF estimate of the conditional law P(Y|X=x) converges to the finite-population and super-population targets in maximum mean discrepancy (MMD), and any continuous functional of the law—quantiles, tolerance regions, covariances—inherits that consistency. If correct, these are the first model-free distributional consistency results for multivariate outcomes under complex surveys, and they provide a principled way to build subgroup-level reference regions from weighted survey data.

What carries the argument

The central object is the survey-weighted kernel mean embedding used both as split criterion and as output: each node's outcome distribution is encoded as a Hájek-weighted average of kernel feature maps, and the MMD between child embeddings guides splits. Two design-aware mechanisms carry the argument: the pseudo-population bootstrap generates multipliers that re-apply the original sampling design, and PSU-level honesty separates split data from estimation data to preserve conditional independence under clustering. The proof's load-bearing identity is a decomposition of the MMD of the forest estimate into five terms whose vanishing is controlled by the design, kernel, and tree-regularity ass

What would settle it

Design a simulation where at some covariate value x the oracle split score has a near-tie between two features and cluster boundaries are placed so that bootstrap resampling flips the chosen split; then the symmetric-difference leaf error will be of order 1/sqrt(N) rather than o_p(d_N), and the MMD between the SDRF estimate and the target conditional law should fail to vanish as N and B grow.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is that a random forest can be rebuilt around the survey design and still provably estimate the full conditional distribution. The estimator replaces the usual mean-squared-error split with the weighted MMD between kernel embeddings of the left and right child distributions, splits at the level of primary sampling units to preserve honesty, and bootstraps by first rebuilding a pseudo-population and then re-applying the sampling design. Theorem 3.4 decomposes the MMD risk into a resampled-leaf term, a resampling-variance term, a design-weight term, an inclusion-design term, and a leaf-approximation term; under the paper's assumptions each term vanishes.

Load-bearing premise

The load-bearing assumption is that the tree leaf built from the bootstrap survey sample converges to the leaf built from the whole finite population, uniformly over query points and faster than the leaf shrinks; the paper calls this hardest to control, and only proves a local single-split version, not the full recursive-tree version.

Editorial extensions

If this is right

  • For survey data, SDRF yields consistent estimates of conditional distributions, not just conditional means, so quantities like conditional quantiles, cross-moments, tail probabilities, and joint tolerance regions can be reported with design-based guarantees.
  • Plug-in functionals inherit consistency under continuity, so downstream analyses such as subgroup comparisons and reference-region construction remain valid under the complex design.
  • Ignoring the design (as in an i.i.d. distributional random forest) leads to bias that does not vanish as N and B grow; simulations under a stratified two-stage PPS design show SDRF reduces RMSE for the conditional mean by roughly half.
  • The MMD-based split criterion is generic over the outcome space, so the same algorithm applies to Euclidean, torus, graph, and simplex-valued outcomes under the stated kernel assumptions.
  • A split-score stability analysis identifies a crossover threshold for averaged resample multipliers; beyond it, deeper trees or larger PSU subsampling fractions matter more than additional resamples.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Editorial extension: the five-term decomposition suggests a concrete diagnostic—track the symmetric-difference leaf error term separately; a design that produces large leaf-boundary instability at deep nodes can break consistency even when marginal splits look stable.
  • Editorial extension: the pseudo-population plus PSU-honesty machinery could plausibly port to other tree ensembles or kernel regressions whose loss is an RKHS distance, since the proof structure does not obviously depend on CART-specific structure beyond split-score Lipschitzness and VC bounds.
  • Editorial extension: the crossover threshold for averaging resample multipliers yields a testable practical rule—averaging more bootstrap replicates at deep nodes will not improve split stability, so computational budget should go to node size or subsampling fraction.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. The paper proposes a survey-calibrated distributional random forest (SDRF) for estimating conditional distributions P(Y|X=x) and continuous functionals of them under complex survey designs. The method combines a pseudo-population bootstrap, PSU-level sample splitting for honesty, and an MMD split criterion computed from Hájek-type design-weighted kernel mean embeddings. The main theoretical contribution is Theorem 3.4, which claims design consistency of the SDRF conditional-distribution estimator in the MMD metric under conditions (D1)-(D4), (R1)-(R3), (S1), (K1)-(K3), a Lipschitz conditional mean embedding, and algorithmic conditions (A1)-(A3), (B1)-(B2); Corollary 3.5 extends this to plug-in functionals. The paper also reports simulations under a stratified two-stage PPS design and a NHANES analysis of conditional tolerance regions for two diabetes biomarkers. The algorithmic design is thoughtful and the simulation results are encouraging, but the central consistency theorem has load-bearing gaps that prevent the stated claims from being accepted as proved.

Significance. If the theoretical claims were valid, this would be a substantial contribution: distributional regression for complex survey designs is an important and underdeveloped area, and the proposed synthesis of pseudo-population bootstrap, PSU-level honesty, and MMD splitting is methodologically natural. The simulation study shows a large and persistent RMSE reduction relative to the unweighted DRF baseline under a stratified two-stage cluster design, and the NHANES illustration is clinically relevant. The decomposition of the MMD risk into design, resampling, and oracle-leaf terms is a useful organizing framework, and Proposition 3.3's analysis of multiplier averaging is informative. However, the main theorem as stated is not established: the finite-population target is not well-defined for continuous covariates, and the pivotal algorithmic assumption (B2) is unproved and is close in strength to the desired conclusion. These are not presentation issues; they affect the central theoretical deliverable.

major comments (3)
  1. [Section 3.3, Theorem 3.4; S7 Part E] The theorem claims d_k(\hat P_{S_N|\omega}(x), P_{S_N|\omega}(x)) \to 0 in design probability. In S7 Part E the target is taken to be \mu_N(x) = (1/N_x)\sum_{i:X_i=x} k(Y_i,\cdot), and the proof asserts that N_x \to \infty with high probability and applies a law of large numbers. For continuous X, P(X_i=x)=0, so N_x is typically 0 or 1 and does not diverge. Even when N_x=1, \mu_N(x) is the embedding of a single observation, and the leaf-averaged estimator T_{L,b}(x) contains an irreducible individual-noise term; the E_N term cannot vanish. Thus design consistency to the exact point-level finite-population conditional law is not established and, as stated, is impossible for continuous covariates. The theorem needs to be reformulated with a well-defined target, e.g., the super-population conditional law or a finite-population target defined through shrinking leaf or kernel neighborhoods, a
  2. [Section 3.3, assumption (B2); S7 Part A] Assumption (B2) requires sup_x P(X \in L^*(x) \Delta L(x)) = o_p(d_N) with d_N = o(k_N^p). This is a global, uniform statement that the data-adaptive leaf L^*(x) built from resampled survey data converges to the oracle leaf L(x) built from the full finite population, at a rate faster than the oracle leaf shrinks. It is used directly in Part A of the proof of Theorem 3.4 and in Lemma S10 to show that the leaf-mismatch term A_N and the 'good tree' sets behave correctly. The paper's own remark after Theorem 3.4 calls (B2) the hardest condition to control, and Theorem 3.2 only establishes a local split-score approximation at a fixed parent node under a unique-maximizer condition. It does not control accumulation of split errors over a recursively grown tree or provide a uniform bound over x of the symmetric difference L^*(x) \Delta L(x). Consequently, the main theorem is conditional on an un
  3. [S7 Part C] In the proof of Part C of Theorem 3.4, the text invokes 'regularity condition (A4)' to control E[(1-I_b)\|T_{L,\nu}\|^3|L(x)] and to justify boundedness of terms in the remainder R_2. However, the assumptions of Theorem 3.4 list only (A1)-(A3); no condition (A4) is defined anywhere in the paper. This is a missing assumption in a step that is needed to bound the resampling-to-design term C_N. The authors should either add and state (A4) as a formal assumption, or prove the required bound from the existing assumptions. As written, the proof has an unsupported step that is load-bearing for the main decomposition.
minor comments (6)
  1. [S7 proof of Theorem 3.4] The definitions of B_L(x) and B_{L^*}(x) in the proof appear garbled: two different sets are both denoted B(x), and the displayed conditions do not consistently distinguish between L(x) and L^*(x). Please rewrite these definitions clearly.
  2. [Algorithm 1] Line 6 uses a 1/q weight for units in D^*_{split}, while line 11 uses 1/(1-q) for the estimation stage. This is presumably correct under PSU-level Bernoulli splitting, but the text should state explicitly that estimation weights are conditional on membership in D^*_{b,est} and define I(D^*_{b,est}) before it is used.
  3. [Proposition 3.3 and Lemma S6] The proof of Proposition 3.3 relies on 'negatively dependent' multipliers and sub-gamma concentration of \bar\delta, but Lemma S6 establishes only pairwise nonpositive covariance and marginal sub-gamma behavior. If full negative association is required for the stated concentration, it should be stated and proved; otherwise the argument should be adjusted.
  4. [Section 2.1] The notation P^N_{Y|X\in A(x)}(x) and P^N_{Y|X=x} is used inconsistently. Since the paper allows continuous X, the distinction between a leaf-region conditional law and a point-level conditional law is not cosmetic; the target measure should be defined precisely in the statement of Theorem 3.4.
  5. [Tables 1-2] Table 1 reports mean (SD) of MMD, but Table 2 reports RMSE without a standard error or interval across the 200 seeds. Adding uncertainty measures would make the simulation comparison more complete.
  6. [Data availability] The code is 'available from the authors upon request.' Given the paper's emphasis on reproducibility, a public repository would be preferable.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: consistency theorem is conditional on an explicit, unproved leaf-convergence assumption (B2), which is a soundness gap rather than a circular reuse of the conclusion.

full rationale

I walked the derivation chain of Theorem 3.4. The MMD consistency proof decomposes the error into terms A_N through E_N. Parts C, D, and E rely on standard design-based law-of-large-numbers/CLT facts, resampling properties (R1)-(R3), and the Lipschitz embedding plus shrinking-diameter condition (B1). Part A is controlled by assumption (B2), the symmetric-difference leaf convergence condition. The paper explicitly flags this: 'Among all the assumptions, (B2) is the hardest to directly control' (Section 3.3, remark after Theorem 3.4). This is an honest statement of a limitation. The assumption is not equivalent to the theorem's conclusion: (B2) concerns sup_x P(X in L*(x) Δ L(x)) = o_p(d_N) with d_N = o(k_N^p), whereas the target is d_k(P_hat, P_SN|ω) → 0. It is a sufficient regularity condition on the algorithmically grown partitions, not a restatement of the target metric. Theorem 3.2 provides only local split consistency; the global uniform leaf convergence needed by (B2) is not proved. This is a genuine soundness/completeness gap, but it is not circularity: the paper does not define its estimator in terms of the target, nor does it fit a parameter to the target and then call it a prediction. The bandwidth is set by the median heuristic and tree hyperparameters are user choices; none enter the asymptotic statement. Self-citations (e.g., Kosorok's empirical-process monograph and Matabuena et al. references) are standard textbook/algorithmic sources with independent content, not load-bearing premises imported to force the conclusion. The pseudo-population bootstrap is attributed to Wang, Peng, and Kim [97], not to the present authors. No step in the paper reduces by construction to its own inputs; score 0.

Assumptions & free parameters 2 free parameters · 5 assumptions · 0 invented entities

The central claim rests on standard survey-design, kernel, and smoothness assumptions plus one unproved algorithmic condition (B2). The method introduces no new physical entities. Kernel bandwidth and tree hyperparameters affect finite samples but are not part of the asymptotic claim.

free parameters (2)
  • Kernel bandwidth (RBF σ) = median of pairwise Euclidean distances between observed Y's (simulations and NHANES)
    Chosen adaptively by the median heuristic; finite-sample SDRF estimates depend on it, though the asymptotic theory only requires a c0-universal kernel.
  • Algorithm hyperparameters: q, max depth, min node size, λmax, mtry, B = In simulations: depth=8; min node size=max(20, √n_s); λmax=5.5·max_i w_i / min_i w_i; mtry=p; B∈{10,30,70,200}; q left u
    User-chosen algorithmic settings that affect finite-sample performance and are intended to enforce regularity conditions (A1)-(A3). They are not fitted to the target distribution.
assumptions (5)
  • domain assumption Design assumptions (D1)-(D4): conditionally noninformative design, stable sampling fraction, bounded first-order inclusion probabilities, controlled second-order inclusion dependence
    Invoked in Lemma 2.1 and throughout Theorems 3.2 and 3.4. Standard in survey sampling, but they exclude designs with extreme weights or heavy second-order dependence.
  • domain assumption Resampling multiplier assumptions (R1)-(R3): design-dependent conditional law, mean 1, bounded moments, weak cross-unit dependence
    Required for the pseudo-population bootstrap to behave. Verified in Lemma 3.1 and the supplement for Poisson, SRSWOR, PPSWR, and stratified two-stage designs.
  • standard math Y is a locally compact Polish space and k is bounded, C0-continuous and c0-universal, so d_k metrizes weak convergence
    Used to identify probability measures through kernel mean embeddings and to convert MMD convergence into weak convergence. Invoked in Sections 2.2 and 3.3.
  • domain assumption Lipschitz conditional mean embedding: ||E[k(Y,·)|X=x] - E[k(Y,·)|X=x']|| ≤ L_lip ||x-x'||
    Needed in Part E of Theorem 3.4's proof to bound the bias from replacing x by a leaf. This is a genuine smoothness condition on the target embedding.
  • ad hoc to paper Algorithmic behavior (A1)-(A3) and (B1)-(B2), especially (B2): sup_x P(X∈L*(x) Δ L(x)) = o_p(d_N)
    The consistency theorem assumes the resampled leaf approximates the oracle leaf uniformly. This is close to the main algorithmic convergence statement and is not proved globally; Theorem 3.2 gives only local split consistency. The paper itself flags (B2) as the hardest condition to control.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Distributional Random Forests for Complex Survey Designs." pith.science (2026). https://pith.science/paper/LAESRMOO

@misc{pith2026251208179,
  author       = {Pith},
  title        = {Pith review of: Distributional Random Forests for Complex Survey Designs},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LAESRMOO}},
  note         = {Machine review of arXiv:2512.08179}
}
abstract

We study estimation of the conditional law $P(Y|X = x)$ and continuous measurable maps of it when $Y \in \mathcal{Y}$ takes values in a locally compact Polish space (e.g., $\mathbb{R}^d$), $X \in \mathbb{R}^p$, and the observations arise from a complex survey design: a single- or multi-stage sampling scheme that may involve unequal selection, stratification, and clustering. We propose a survey-calibrated distributional random forest (SDRF) that incorporates complex-design features via the pseudo-population bootstrap, PSU-level honesty, and a Maximum Mean Discrepancy (MMD) split criterion computed from kernel mean embeddings of design-weighted node distributions. We provide a framework for analyzing forest-based estimators under various survey designs; establish consistency for both finite- and super-population conditional laws 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 expose the systematic bias incurred by ignoring survey structure. We illustrate the broad applicability of SDRF on NHANES, estimating the conditional joint tolerance regions for two diabetes biomarkers, revealing subgroup-level distributional heterogeneity relevant to diabetes risk profiling in the U.S. population.

Figures

Figures reproduced from arXiv: 2512.08179 by the authors.

Figure 1
Figure 1. Pointwise mean square error (MSE) of the estimator [PITH_FULL_IMAGE:figures/full_fig_p019_1.png] view at source ↗
Figure 2
Figure 2. Pointwise standard error (SD) of the estimator [PITH_FULL_IMAGE:figures/full_fig_p019_2.png] view at source ↗
Figure 3
Figure 3. SDRF–estimated tolerance regions (in expectation) for [PITH_FULL_IMAGE:figures/full_fig_p022_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: SDRF–estimated tolerance regions (in expectation) for [PITH_FULL_IMAGE:figures/full_fig_p023_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

106 extracted references · 3 linked inside Pith

  1. [1]

    Theory of reproducing kernels.Transactions of the American Mathematical Society, 68(3):337–404, 1950

    Nachman Aronszajn. Theory of reproducing kernels.Transactions of the American Mathematical Society, 68(3):337–404, 1950

  2. [2]

    American Diabetes Association. 2. diagnosis and classification of diabetes: Standards of care in diabetes—2025.Diabetes Care, 48(Supplement_1):S27–S49, 12 2024

  3. [3]

    Recursive partitioning for heterogeneous causal effects

    Susan Athey and Guido Imbens. Recursive partitioning for heterogeneous causal effects. Proc. Natl. Acad. Sci. U. S. A., 113:7353–7360, 2016

  4. [4]

    Generalized random forests.Ann

    Susan Athey, Julie Tibshirani, and Stefan Wager. Generalized random forests.Ann. Stat., 47:1148–1178, 2019

  5. [5]

    On deep learning as a remedy for the curse of dimensionality in nonparametric regression.The Annals of Statistics, 47(4):2261–2285, 8 2019

    Benedikt Bauer and Michael Kohler. On deep learning as a remedy for the curse of dimensionality in nonparametric regression.The Annals of Statistics, 47(4):2261–2285, 8 2019

  6. [6]

    The central limit theorem under simple random sampling.The American Statistician, 55(4):352–357, 2001

    David R Bellhouse. The central limit theorem under simple random sampling.The American Statistician, 55(4):352–357, 2001

  7. [7]

    Mmd-based variable importance for distribu- tional random forest

    Clément Bénard, Jeffrey Näf, and Julie Josse. Mmd-based variable importance for distribu- tional random forest. InInternational Conference on Artificial Intelligence and Statistics, pages 1324–1332. PMLR, 2024. 24

  8. [8]

    Is now the time to review the diabetes diagnostic criteria?Diabetes/Metabolism Research and Reviews, 41(4):e70046, 2025

    Maja Cigrovski Berkovic, Felice Strollo, and Antonio Ceriello. Is now the time to review the diabetes diagnostic criteria?Diabetes/Metabolism Research and Reviews, 41(4):e70046, 2025

Show all 106 references
  1. [9]

    Consistency of random forests and other averaging classifiers.Journal of Machine Learning Research, 9(9), 2008

    Gérard Biau, Luc Devroye, and Gäbor Lugosi. Consistency of random forests and other averaging classifiers.Journal of Machine Learning Research, 9(9), 2008

  2. [10]

    A random forest guided tour.Test, 25(2):197–227, 2016

    Gérard Biau and Erwan Scornet. A random forest guided tour.Test, 25(2):197–227, 2016

  3. [11]

    Functional central limit theorems for single-stage sampling designs.Annals of Statistics, 2017

    Hélène Boistard, Hendrik P Lopuhaä, and Anne Ruiz-Gazen. Functional central limit theorems for single-stage sampling designs.Annals of Statistics, 2017

  4. [12]

    Approximation of rejective sampling inclusion probabilities and application to high order correlations.arXiv [math.ST], 2012

    Hélène Boistard, Hendrik P Lopuhaä, and Anne Ruiz-Gazen. Approximation of rejective sampling inclusion probabilities and application to high order correlations.arXiv [math.ST], 2012

  5. [13]

    Lopuhaä, and Anne Ruiz-Gazen

    Hélène Boistard, Hendrik P. Lopuhaä, and Anne Ruiz-Gazen. Functional central limit theorems for single-stage sampling designs.The Annals of Statistics, 45(4), 2017

  6. [14]

    Model-assisted estimation for complex surveys using penalised splines.Biometrika, 92(4):831–846, 2005

    F Jay Breidt, Gerda Claeskens, and JD Opsomer. Model-assisted estimation for complex surveys using penalised splines.Biometrika, 92(4):831–846, 2005

  7. [15]

    Bagging predictors.Machine learning, 24(2):123–140, 1996

    Leo Breiman. Bagging predictors.Machine learning, 24(2):123–140, 1996

  8. [16]

    Random forests.Machine learning, 45:5–32, 2001

    Leo Breiman. Random forests.Machine learning, 45:5–32, 2001

  9. [17]

    Analyzing bagging.Ann

    Peter Bühlmann and Bin Yu. Analyzing bagging.Ann. Stat., 30:927–961, 2002

  10. [18]

    Properties of design-based functional principal components analysis.J

    Hervé Cardot, Mohamed Chaouch, Camelia Goga, and Catherine Labruère. Properties of design-based functional principal components analysis.J. Stat. Plan. Inference, 140:75–91, 2010

  11. [19]

    Vector valued reproducing kernel hilbert spaces and universality.Anal

    C Carmeli, E de Vito, A Toigo, and V Umanità. Vector valued reproducing kernel hilbert spaces and universality.Anal. Appl., 08:19–61, 2010

  12. [20]

    Distributional random forests: Heterogeneity adjustment and multivariate distributional regression.Journal of Machine Learning Research, 23, 2022

    Domagoj Cevid, Loris Michel, N Meinshausen, and P Bühlmann. Distributional random forests: Heterogeneity adjustment and multivariate distributional regression.Journal of Machine Learning Research, 23, 2022

  13. [21]

    Challenges in statistics: A dozen challenges in causality and causal inference, 2025

    Carlos Cinelli, Avi Feller, Guido Imbens, Edward Kennedy, Sara Magliacane, and Jose Zubizarreta. Challenges in statistics: A dozen challenges in causality and causal inference, 2025

  14. [22]

    A unified principled framework for resampling based on pseudo-populations: Asymptotic theory

    Pier Luigi Conti, Daniela Marella, Fulvia Mecatti, and Federico Andreis. A unified principled framework for resampling based on pseudo-populations: Asymptotic theory. Bernoulli (Andover.), 26:1044–1069, 2020

  15. [23]

    Resampling under complex sampling designs: Roots, development and the way forward.Stats, 5(1):258–269, 2022

    Pier Luigi Conti and Fulvia Mecatti. Resampling under complex sampling designs: Roots, development and the way forward.Stats, 5(1):258–269, 2022. 25

  16. [24]

    Chapman and Hall/CRC, 2024

    Ciprian M Crainiceanu, Jeff Goldsmith, Andrew Leroux, and Erjia Cui.Functional data analysis with R. Chapman and Hall/CRC, 2024

  17. [25]

    Springer Berlin Heidelberg, Berlin, Heidelberg, 2025

    Jill DeMatteis and Graham Kalton.Sampling from Finite Populations, pages 2247–2253. Springer Berlin Heidelberg, Berlin, Heidelberg, 2025

  18. [26]

    Gene selection and classification of microarray data using random forest.BMC bioinformatics, 7(1):3, 2006

    Ramón Díaz-Uriarte and Sara Alvarez de Andrés. Gene selection and classification of microarray data using random forest.BMC bioinformatics, 7(1):3, 2006

  19. [27]

    Cart and best-ortho-basis: a connection.The Annals of statistics, 25(5):1870–1911, 1997

    David L Donoho. Cart and best-ortho-basis: a connection.The Annals of statistics, 25(5):1870–1911, 1997

  20. [28]

    Jaroslav hajek, 1926-1974, 1975

    Václav Dupač. Jaroslav hajek, 1926-1974, 1975

  21. [29]

    Monographs on Statistics and Applied Probability (Closed)

    Bradley Efron and Robert Tibshirani.An introduction to the bootstrap. Monographs on Statistics and Applied Probability (Closed). Springer, 1993

  22. [30]

    Subjective bayesian models in sampling finite populations.Journal of the Royal Statistical Society Series B: Statistical Methodology, 31(2):195–224, 1969

    William A Ericson. Subjective bayesian models in sampling finite populations.Journal of the Royal Statistical Society Series B: Statistical Methodology, 31(2):195–224, 1969

  23. [31]

    Do we need hundreds of classifiers to solve real world classification problems?The journal of machine learning research, 15(1):3133–3181, 2014

    Manuel Fernández-Delgado, Eva Cernadas, Senén Barro, and Dinani Amorim. Do we need hundreds of classifiers to solve real world classification problems?The journal of machine learning research, 15(1):3133–3181, 2014

  24. [32]

    On the rate of convergence in wasserstein distance of the empirical measure.Probab

    Nicolas Fournier and Arnaud Guillin. On the rate of convergence in wasserstein distance of the empirical measure.Probab. Theory Relat. Fields, 162:707–738, 2015

  25. [33]

    Kernel k-groups via hartigan’s method.IEEE Transactions on Pattern Analysis and Machine Intelligence, 43(12):4411– 4425, 2020

    Guilherme França, Maria L Rizzo, and Joshua T Vogelstein. Kernel k-groups via hartigan’s method.IEEE Transactions on Pattern Analysis and Machine Intelligence, 43(12):4411– 4425, 2020

  26. [34]

    Nonparametric tolerance regions.The Annals of Mathematical Statistics, 24(1):44–55, 1953

    Donald AS Fraser. Nonparametric tolerance regions.The Annals of Mathematical Statistics, 24(1):44–55, 1953

  27. [35]

    Asymptotic normality of the median heuristic.arXiv [math.ST], 7 2017

    Damien Garreau. Asymptotic normality of the median heuristic.arXiv [math.ST], 7 2017

  28. [36]

    Random forests for time-dependent processes.ESAIM: Probability and Statistics, 24:801–826, 2020

    Benjamin Goehry. Random forests for time-dependent processes.ESAIM: Probability and Statistics, 24:801–826, 2020

  29. [37]

    Akernel two-sampletest.The Journal of Machine Learning Research, 13(1):723–773, 2012

    Arthur Gretton, Karsten M Borgwardt, Malte J Rasch, Bernhard Schölkopf, and Alexander Smola. Akernel two-sampletest.The Journal of Machine Learning Research, 13(1):723–773, 2012

  30. [38]

    Borgwardt, Malte J

    Arthur Gretton, Karsten M. Borgwardt, Malte J. Rasch, Bernhard Schölkopf, and Alexan- der J. Smola. A kernel method for the two-sample problem.CoRR, abs/0805.2368, 2008

  31. [39]

    Optimal strategy and other problems in probability sampling.Časopis pro pěstování matematiky, 84(4):387–423, 1959

    Jaroslav Hájek. Optimal strategy and other problems in probability sampling.Časopis pro pěstování matematiky, 84(4):387–423, 1959. 26

  32. [40]

    Limiting distributions in simple random sampling from a finite population

    Jaroslav Hájek. Limiting distributions in simple random sampling from a finite population. A Magyar Tudományos Akadémia Matematikai Kutató Intézetének közlemenyei, 5(3):361– 374, 1960

  33. [41]

    Qiyang Han and Jon A. Wellner. Convergence rates of least squares regression estimators with heavy-tailed errors.The Annals of Statistics, 47(4):2286–2319, 2019

  34. [42]

    Qiyang Han and Jon A. Wellner. Complex sampling designs: Uniform limit theorems and applications.The Annals of Statistics, 49(1):459 – 485, 2021

  35. [43]

    On kernel nonparametric regression designed for complex survey data.Metrika, 72(1):111–138, 2010

    Torsten Harms and Pierre Duchesne. On kernel nonparametric regression designed for complex survey data.Metrika, 72(1):111–138, 2010

  36. [44]

    A generalization of sampling without replacement from a finite universe.Journal of the American statistical Association, 47(260):663–685, 1952

    Daniel G Horvitz and Donovan J Thompson. A generalization of sampling without replacement from a finite universe.Journal of the American statistical Association, 47(260):663–685, 1952

  37. [45]

    Cambridge university press, 2015

    Guido W Imbens and Donald B Rubin.Causal inference in statistics, social, and biomedical sciences. Cambridge university press, 2015

  38. [46]

    Variable selection with lasso regression for complex survey data.Stat, 12(1):e578, 2023

    Amaia Iparragirre, Thomas Lumley, Irantzu Barrio, and Inmaculada Arostegui. Variable selection with lasso regression for complex survey data.Stat, 12(1):e578, 2023

  39. [47]

    On the estimation of the characteristic function in finite populations with applications.Test, 27(1):95–121, 2018

    MD Jiménez-Gamero, JL Moreno-Rebollo, and JA Mayor-Gallego. On the estimation of the characteristic function in finite populations with applications.Test, 27(1):95–121, 2018

  40. [48]

    On local linear regression estimation in sampling surveys.Journal Name, 2017

    Conlet B Kikechi, Richard O Simwa, and Ganesh P Pokhariyal. On local linear regression estimation in sampling surveys.Journal Name, 2017

  41. [49]

    Wiley Classics Library

    Leslie Kish.Survey Sampling. Wiley Classics Library. John Wiley & Sons, 1995

  42. [50]

    Rage against the mean–a review of distributional regression approaches.Econometrics and Statistics, 26:99–123, 2023

    Thomas Kneib, Alexander Silbersdorff, and Benjamin Säfken. Rage against the mean–a review of distributional regression approaches.Econometrics and Statistics, 26:99–123, 2023

  43. [51]

    Strategies in the multi- variate analysis of data from complex surveys.International Statistical Review/Revue Internationale de Statistique, pages 59–78, 1975

    Gary G Koch, Daniel H Freeman Jr, and Jean L Freeman. Strategies in the multi- variate analysis of data from complex surveys.International Statistical Review/Revue Internationale de Statistique, pages 59–78, 1975

  44. [52]

    Function on scalar regression with complex survey designs, 2025

    Lily Koffman, Sunan Gao, Xinkai Zhou, Andrew Leroux, Ciprian Crainiceanu, and John Muschelli III. Function on scalar regression with complex survey designs, 2025

  45. [53]

    Springer, 2008

    Michael R Kosorok.Introduction to empirical inference processes and semiparametric inference. Springer, 2008

  46. [54]

    Modeling conditional reference regions: Application to glycemic markers.Statistics in Medicine, 40(26):5926–5946, 2021

    Óscar Lado-Baleato, Javier Roca-Pardiñas, Carmen Cadarso-Suárez, and Francisco Gude. Modeling conditional reference regions: Application to glycemic markers.Statistics in Medicine, 40(26):5926–5946, 2021. 27

  47. [55]

    Random forests and adaptive nearest neighbors.Journal of the American Statistical Association, 101(474):578–590, 2006

    Yi Lin and Yongho Jeon. Random forests and adaptive nearest neighbors.Journal of the American Statistical Association, 101(474):578–590, 2006

  48. [56]

    Randomization can reduce both bias and variance: A case study in random forests.Journal of Machine Learning Research, 26(150):1–49, 2025

    Brian Liu and Rahul Mazumder. Randomization can reduce both bias and variance: A case study in random forests.Journal of Machine Learning Research, 26(150):1–49, 2025

  49. [57]

    Chapman & Hall/CRC, 3 edition, 2021

    Sharon L Lohr.Sampling: Design and analysis. Chapman & Hall/CRC, 3 edition, 2021

  50. [58]

    Conformal and knn predictive uncertainty quantifi- cation algorithms in metric spaces.arXiv preprint arXiv:2507.15741, 2025

    Gábor Lugosi and Marcos Matabuena. Conformal and knn predictive uncertainty quantifi- cation algorithms in metric spaces.arXiv preprint arXiv:2507.15741, 2025

  51. [59]

    Conformal and knn predictive uncertainty quantifi- cation algorithms in metric spaces, 2025

    Gábor Lugosi and Marcos Matabuena. Conformal and knn predictive uncertainty quantifi- cation algorithms in metric spaces, 2025

  52. [60]

    John Wiley & Sons, 2011

    Thomas Lumley.Complex surveys: a guide to analysis using R. John Wiley & Sons, 2011

  53. [61]

    An empirical-process central limit theorem for complex sampling under bounds on the design effect, 2012

    Thomas Lumley. An empirical-process central limit theorem for complex sampling under bounds on the design effect, 2012

  54. [62]

    Fitting regression models to survey data.Statistical Science, pages 265–278, 2017

    Thomas Lumley and Alastair Scott. Fitting regression models to survey data.Statistical Science, pages 265–278, 2017

  55. [63]

    Strong negative type in spheres.Pacific Journal of Mathematics, 307(2):383– 390, 2020

    Russell Lyons. Strong negative type in spheres.Pacific Journal of Mathematics, 307(2):383– 390, 2020

  56. [64]

    A survey of bootstrap methods in finite population sampling.Stat

    Zeinab Mashreghi, David Haziza, and Christian Léger. A survey of bootstrap methods in finite population sampling.Stat. Surv., 10:1–52, 2016

  57. [65]

    Hy- pothesis testing for matched pairs with missing data by maximum mean discrepancy: An application to continuous glucose monitoring.The American Statistician, 77(4):357–369, 2023

    Marcos Matabuena, Paulo Félix, Marc Ditzhaus, Juan Vidal, and Francisco Gude. Hy- pothesis testing for matched pairs with missing data by maximum mean discrepancy: An application to continuous glucose monitoring.The American Statistician, 77(4):357–369, 2023

  58. [66]

    Marcos Matabuena, Aritra Ghosal, Wendy Meiring, and Alexander Petersen. Predicting distributions of physical activity profiles in the national health and nutrition examina- tion survey database using a partially linear fréchet single index model.Biostatistics, 26(1):kxaf013, 05 2025

  59. [67]

    Vidal, Rahul Ghosal, and Jukka-Pekka Onnela

    Marcos Matabuena, Juan C. Vidal, Rahul Ghosal, and Jukka-Pekka Onnela. Screening for diabetes mellitus in the u.s. population using neural network models and complex survey designs, 2025

  60. [68]

    Ker- nel biclustering algorithm in hilbert spaces.Advances in Data Analysis and Classification, pages 1–42, 2025

    Marcos Matabuena, Juan C Vidal, Oscar Hernan Madrid Padilla, and Dino Sejdinovic. Ker- nel biclustering algorithm in hilbert spaces.Advances in Data Analysis and Classification, pages 1–42, 2025

  61. [69]

    Persistent increase of prevalence of metabolic syndrome among us adults: Nhanes iii to nhanes 1999–2006.Diabetes care, 34(1):216–219, 2011

    Arupendra Mozumdar and Gary Liguori. Persistent increase of prevalence of metabolic syndrome among us adults: Nhanes iii to nhanes 1999–2006.Diabetes care, 34(1):216–219, 2011. 28

  62. [70]

    A general semiparametric z-estimation approach for case- cohort studies.Statistica Sinica, 23(3):1155, 2013

    Bin Nan and Jon A Wellner. A general semiparametric z-estimation approach for case- cohort studies.Statistica Sinica, 23(3):1155, 2013

  63. [71]

    An extension of the wilcoxon rank sum test for complex sample survey data.Journal of the Royal Statistical Society Series C: Applied Statistics, 61(4):653–664, 2012

    Sundar Natarajan, Stuart R Lipsitz, Garrett M Fitzmaurice, Debajyoti Sinha, Joseph G Ibrahim, Jennifer Haas, and Walid Gellad. An extension of the wilcoxon rank sum test for complex sample survey data.Journal of the Royal Statistical Society Series C: Applied Statistics, 61(4)...

  64. [72]

    The reproducibility crisis in science: A statistical counterattack.Significance, 12(3):30–32, 2015

    Roger Peng. The reproducibility crisis in science: A statistical counterattack.Significance, 12(3):30–32, 2015

  65. [73]

    Random forest weighted local fréchet regression with random objects.Journal of Machine Learning Research, 25(107):1–69, 2024

    Rui Qiu, Zhou Yu, and Ruoqing Zhu. Random forest weighted local fréchet regression with random objects.Journal of Machine Learning Research, 25(107):1–69, 2024

  66. [74]

    R Foundation for Statistical Computing, Vienna, Austria, 2024

    R Core Team.R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria, 2024

  67. [75]

    Multilevel modelling of complex survey data

    Sophia Rabe-Hesketh and Anders Skrondal. Multilevel modelling of complex survey data. Journal of the Royal Statistical Society Series A: Statistics in Society, 169(4):805–827, 2006

  68. [76]

    On simple adjustments to chi-square tests with sample survey data.The annals of statistics, pages 385–397, 1987

    JNK Rao and AJ Scott. On simple adjustments to chi-square tests with sample survey data.The annals of statistics, pages 385–397, 1987

  69. [77]

    Resampling inference with complex survey data.Journal of the american statistical association, 83(401):231–241, 1988

    Jon NK Rao and CFJ Wu. Resampling inference with complex survey data.Journal of the american statistical association, 83(401):231–241, 1988

  70. [78]

    On the subspaces ofl p (p>2) spanned by sequences of independent random variables.Israel Journal of Mathematics, 8(3):273–303, 9 1970

    Haskell P Rosenthal. On the subspaces ofl p (p>2) spanned by sequences of independent random variables.Israel Journal of Mathematics, 8(3):273–303, 9 1970

  71. [79]

    Estimating the mean of a finite population.The Annals of Mathematical Statistics, pages 392–398, 1960

    Jogabrata Roy and IM Chakravarti. Estimating the mean of a finite population.The Annals of Mathematical Statistics, pages 392–398, 1960

  72. [80]

    Walter Rudin.Fourier Analysis on Groups. Wiley. eBook edition; originally published 1962

  73. [81]

    Design-based ratio estimators and central limit theorems for clustered, blocked rcts.Journal of the American Statistical Association, 117(540):2135–2146, 2022

    Peter Z Schochet, Nicole E Pashley, Luke W Miratrix, and Tim Kautz. Design-based ratio estimators and central limit theorems for clustered, blocked rcts.Journal of the American Statistical Association, 117(540):2135–2146, 2022

  74. [82]

    Random forests and kernel methods.IEEE Transactions on Information Theory, 62(3):1485–1500, 2016

    Erwan Scornet. Random forests and kernel methods.IEEE Transactions on Information Theory, 62(3):1485–1500, 2016

  75. [83]

    Equiva- lence of distance-based and RKHS-based statistics in hypothesis testing.The Annals of Statistics, 41(5), 2013

    Dino Sejdinovic, Bharath Sriperumbudur, Arthur Gretton, and Kenji Fukumizu. Equiva- lence of distance-based and RKHS-based statistics in hypothesis testing.The Annals of Statistics, 41(5), 2013

  76. [84]

    Population and sampling.JAMA, 186(5):499–502, 1963

    Sidney Shindell. Population and sampling.JAMA, 186(5):499–502, 1963. 29

  77. [85]

    Using lasso to model interactions and nonlinearities in survey data.Survey Practice, 11(1), 2018

    Curtis S Signorino and Antje Kirchner. Using lasso to model interactions and nonlinearities in survey data.Survey Practice, 11(1), 2018

  78. [86]

    A hilbert space embedding for distributions

    Alex Smola, Arthur Gretton, Le Song, and Bernhard Scholkopf. A hilbert space embedding for distributions. InIFIP Working Conference on Database Semantics, 2007

  79. [87]

    Universality, characteristic kernels and RKHS embedding of measures.arXiv [stat.ML], 2010

    Bharath K Sriperumbudur, Kenji Fukumizu, and Gert R G Lanckriet. Universality, characteristic kernels and RKHS embedding of measures.arXiv [stat.ML], 2010

  80. [88]

    Sriperumbudur, Arthur Gretton, Kenji Fukumizu, Bernhard Schölkopf, and Gert R.G

    Bharath K. Sriperumbudur, Arthur Gretton, Kenji Fukumizu, Bernhard Schölkopf, and Gert R.G. Lanckriet. Hilbert space embeddings and metrics on probability measures. Journal of Machine Learning Research, 11:1517––1561, 8 2010

  81. [89]

    On the influence of the kernel on the consistency of support vector machines.J

    Ingo Steinwart. On the influence of the kernel on the consistency of support vector machines.J. Mach. Learn. Res., 2:67–93, 2001

  82. [90]

    E-statistics: The energy of statistical samples.Bowling Green State University, Department of Mathematics and Statistics Technical Report, 3(05):1–18, 2003

    Gábor J Székely. E-statistics: The energy of statistical samples.Bowling Green State University, Department of Mathematics and Statistics Technical Report, 3(05):1–18, 2003

  83. [91]

    Springer International Publishing, 2 edition, 2023

    A W van der Vaart and Jon A Wellner.Weak convergence and empirical processes. Springer International Publishing, 2 edition, 2023

  84. [92]

    M-estimators

    Aad W van der Vaart and Jon A Wellner. M-estimators. InWeak Convergence and Empirical Processes: With Applications to Statistics, pages 284–308. Springer, 1996

  85. [93]

    V. S. Varadarajan. On the convergence of sample probability distributions.Sankhy¯ a: The Indian Journal of Statistics, 19(1/2):23–26, 1958

  86. [94]

    Forest guided smoothing.arXiv preprint arXiv:2103.05092, 2021

    Isabella Verdinelli and Larry Wasserman. Forest guided smoothing.arXiv preprint arXiv:2103.05092, 2021

  87. [95]

    Estimation and inference of heterogeneous treatment effects using random forests.J

    Stefan Wager and Susan Athey. Estimation and inference of heterogeneous treatment effects using random forests.J. Am. Stat. Assoc., 113:1228–1242, 2018

  88. [96]

    Adaptive concentration of regression trees, with application to random forests

    Stefan Wager and Guenther Walther. Adaptive concentration of regression trees, with application to random forests. Version Number: 3

  89. [97]

    Bootstrap inference for the finite population mean under complex sampling designs.Journal of the Royal Statistical Society Series B: Statistical Methodology, 84(4):1150–1174, 2022

    Zhonglei Wang, Liuhua Peng, and Jae Kwang Kim. Bootstrap inference for the finite population mean under complex sampling designs.Journal of the Royal Statistical Society Series B: Statistical Methodology, 84(4):1150–1174, 2022

  90. [98]

    Frank Yates.Sampling methods for censuses and surveys.Charles Griffin, 1953

  91. [99]

    Sir ronald fisher and the design of experiments.Biometrics, 20(2):307–321, 1964

    Frank Yates. Sir ronald fisher and the design of experiments.Biometrics, 20(2):307–321, 1964

  92. [100]

    Deep com- positional spatial models.Journal of the American Statistical Association, 117(540):1787– 1808, 2022

    Andrew Zammit-Mangion, Tin Lok James Ng, Quan Vu, and Maurizio Filippone. Deep com- positional spatial models.Journal of the American Statistical Association, 117(540):1787– 1808, 2022. 30 Supplementary Material S1 Notation Table Table S1: Summary of notation Symbol Descriptio...

  93. [101]

    Under (K1), each ofPY|X∈A(x) , the true conditional measure in the super population, PN Y|X∈A(x) , the true conditional measure in the finite population, andˆP HT Y|X∈A(x) and ˆP HJ Y|X∈A(x) , its estimators, is a finite–signed measure inMb(Y)

  94. [102]

    Thus, HT/Hájek properties transfer toH-valued objects by linearity, as the additional randomness enters through from the weights

    For any finite collection{(ai,yi)}with ∑ i|ai|<∞, µk (∑ i aiδyi ) = ∑ i ai k(·,yi),⟨µ k( ∑ i aiδyi),f⟩H = ∑ i aif(yi). Thus, HT/Hájek properties transfer toH-valued objects by linearity, as the additional randomness enters through from the weights. •Dirac measure (at a fixed p...

  95. [103]

    The resamples used to start a tree maintain the original design structure

  96. [104]

    S5.1 Resampling Proof of Lemma 3.1.We focus on proving (R2) and (R3) for some single–stage designs, as (R1) holds directly from the re–sampling procedure detailed below

    Honesty remains in that the data used for the estimation ofˆωi are separate from the data used to determine the splits. S5.1 Resampling Proof of Lemma 3.1.We focus on proving (R2) and (R3) for some single–stage designs, as (R1) holds directly from the re–sampling procedure det...

  97. [105]

    good trees

    Taking expectation overp∗|ξ yields E(n∗ i|ξ) =ξi, which implies the requiredE(n∗ i|ξ ) = 1 +o(1). Consider EN :={f −≤n s/N≤f +}withf − andf + constants, EN satisfiesPSN|ω(EN)→ 1by (D2). OnEN, using (S7): since∑n k=1N∗ k =N andpa(k)≤c +/N, S∗ =∑n k=1N∗ kpa(k)≤∑n k=1N∗ k c+ N =c...

  98. [106]

    Proof of Lemma S10.Part 1: We first control the mean and variance for the oracle leaf–based sizes

    For any tree indexb, PSN|ω(b∈B L(x) ∆BL∗(x)) =o ( 1 Nkp N ) +o ( dN kp N ) , so in particular whend N =o(k p N)andNk p N→∞then PSN|ω(b∈B L(x) ∆BL∗(x)) =o(1). Proof of Lemma S10.Part 1: We first control the mean and variance for the oracle leaf–based sizes. DefineS b :=∑N i=1n∗...

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.