Pith. sign in

REVIEW 3 major objections 3 minor 48 references

Fairness-aware Bayes optimal functional classification

T0 review · 3 major / 3 minor · reviewed 2026-08-15 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read This paper establishes finite-sample guarantees for a fairness-constrained Bayes-optimal functional classifier built from Radon-Nikodym derivatives, and proves a post-processed FLDA variant achieves them.

desk verdict First fairness-aware functional classification framework with a novel FLDA excess-risk byproduct, but the headline fairness guarantee has a proof gap that needs fixing. read the letter →

arxiv 2505.09471 v1 pith:USV2J3YH submitted 2025-05-14 stat.ML cs.LGmath.STstat.MEstat.TH

Xiaoyu Hu , Gengyu Xue , Zhenhua Lin , Yi Yu This is my paper · ORCID
classification stat.MLcs.LGmath.STstat.MEstat.TH MSC 62H3062R1060G1562G20
keywords functionaldataclassificationalgorithmicfairnessBayesoptimalRadon-NikodymderivativeGaussianprocesseslineardiscriminantanalysisexcessriskcontrolpost-processing
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

This paper extends fairness-constrained classification from finite-dimensional features to functional data, where each observation is a curve. It characterizes the fair Bayes-optimal classifier as a group-wise thresholding rule driven by the Radon-Nikodym derivative of the two class-conditional Gaussian processes, then proposes a post-processing estimator, Fair-FLDA, that plugs in truncated spectral estimates and calibrates the threshold on a separate sample. The main theorem states that, with high probability, the calibrated classifier keeps the population disparity below a pre-specified level up to an $O(n^{-1/2})$ offset and keeps its misclassification risk within an explicitly quantified distance from the fair oracle risk. If true, this gives the first provable fairness-accuracy trade-off for functional classification, with the unconstrained FLDA excess-risk bound emerging as a special case.

What carries the argument

The central object is the Radon-Nikodym derivative $\eta_a = dP_{a,1}/dP_{a,0}$ of the two class-conditional Gaussian measures within each sensitive group, which replaces the posterior probability that is intractable in infinite-dimensional spaces. Because misclassification risk and bilinear disparity measures are linear functionals of the randomized classifier, the generalized Neyman-Pearson lemma yields a closed-form thresholded decision rule, with fairness enforced by shifting the threshold through the parameter $\tau$. The finite-sample algorithm estimates $\eta_a$ through a truncated spectral expansion of the covariance, using estimated eigenvalues, eigenfunctions, and principal component scores, and then selects $\hat\tau$ on a holdout set to satisfy the empirical disparity bound. The spectral decay rate $\alpha$ of the eigenvalues and the alignment exponent $\beta$ of the mean difference control the bias-variance trade-off of the truncation level $J$ and drive the convergence rates in the excess-risk bound.

What would settle it

Simulate two Gaussian-process classes whose mean difference lies outside the reproducing kernel Hilbert space of the covariance (for example, a mean difference with Fourier coefficients decaying slower than the square root of the eigenvalues, so perfect classification is possible), run Fair-FLDA at a small disparity level $\delta$, and check whether the empirical disparity violations exceed the theorem's $\delta + O(n^{-1/2})$ bound or whether the excess risk diverges; the theorem is undefined in this regime, so any observed breakdown or surprising stability would directly test the scope of the claim.

Watch

Extended reading notes

Core claim

Within each sensitive group $a$, the two classes are modeled as Gaussian processes $\mathrm{GP}(\mu_{a,y}, K_a)$ that are mutually absolutely continuous, so the likelihood ratio $\eta_a(X)=dP_{a,1}/dP_{a,0}(X)$ exists. For any bilinear disparity measure (equality of opportunity, predictive equality, or demographic parity), the $\delta$-fair Bayes classifier takes the form $g_{D,\tau}(x,a)=1\{(\pi_{a,1}-\tau s_{D,a})\eta_a(x) \ge \pi_{a,0}+\tau b_{D,a}\}$, where $\tau^*_{D,\delta}$ is the smallest magnitude $|\tau|$ for which the disparity of $g_{D,\tau}$ is at most $\delta$. Fair-FLDA estimates $\eta_a$ from a training sample by retaining the first $J$ principal components of an estimated covariance operator, then calibrates $\tau$ on an independent calibration sample to satisfy the empirical disparity constraint. The paper proves that with probability at least $1-\eta$ the population disparity of the output satisfies $|D(\hat f_{D,\delta})|\le \delta + O(\sqrt{\log(1/\eta)/n})$, and the risk gap to the fair oracle satisfies $|R(\hat f_{D,\delta})-R(f^*_{D,\delta})| \lesssim d_E(\hat f_{D,\delta}, f^*_{D,\delta}) + |\tau^*_{D,\delta}|\sqrt{\log(1/\eta)/n}$, with $d_E \lesssim (\epsilon_\pi+\epsilon_\eta+\epsilon_\tau)^2$.

Load-bearing premise

The entire construction presupposes that, within each sensitive group, the two class-conditional Gaussian measures are mutually absolutely continuous, equivalently that the mean difference lies in the reproducing kernel Hilbert space of the covariance; this rules out perfect classification, and when it fails the Radon-Nikodym derivative does not exist, so the plug-in framework and its guarantees no longer apply.

Editorial extensions

If this is right

  • If the theorem is correct, a practitioner can pre-specify a disparity tolerance $\delta$ and obtain the guarantee that the population disparity stays below $\delta + O(n^{-1/2})$ with high probability; using a slightly smaller calibration target $\delta - C\sqrt{\log(1/\eta)/n}$ gives population-level control below $\delta$.
  • The fairness-accuracy trade-off is made explicit: tightening $\delta$ increases $|\tau^*_{D,\delta}|$, which adds a cost of order $|\tau^*_{D,\delta}|\sqrt{\log(1/\eta)/n}$ to the excess-risk bound, and $\epsilon_\tau$, the estimation error of the threshold, enters quadratically in the fairness-aware excess risk.
  • The unconstrained functional linear discriminant analysis excess-risk bound, with unknown eigenfunctions, follows as the special case $\tau^*=0$, which is the first time such a rate has been established without assuming known eigenfunctions.
  • The optimal truncation level $J$ is determined by the eigenvalue decay rate $\alpha$ and the mean-alignment exponent $\beta$, giving an explicit rule for how many principal components to retain in practice.

Reading between the lines

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

  • The authors leave implicit that the same thresholding characterization should extend to heteroscedastic Gaussian processes, since a Gaussian likelihood ratio remains explicit when $K_{a,0}\neq K_{a,1}$; the spectral estimation would simply become class- and group-specific.
  • A testable extension is the perfect-classification regime that the paper excludes: when $\mu_{a,1}-\mu_{a,0}\notin H(K_a)$, the density ratio fails to exist, yet the paper's Remark 2 suggests automatic fairness for equality of opportunity and predictive equality; a regularized surrogate for the density ratio could allow the same risk control to be studied there.
  • The parameter $|\tau^*_{D,\delta}|$ functions as an interpretable fairness-accuracy cost coefficient, and estimating it alongside a deployed classifier would give stakeholders a concrete number describing how much accuracy is being sacrificed to meet the disparity constraint.
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 / 3 minor

Summary. The paper develops a framework for Bayes-optimal binary classification of functional data under a linear disparity constraint. It first characterizes the population fair classifier through Radon–Nikodym derivatives and the generalized Neyman–Pearson lemma (Theorem 2), then specializes to homoscedastic Gaussian processes and proposes Fair-FLDA (Algorithm 1), a plug-in classifier with estimated means, covariances, eigenvalues/eigenfunctions, and a calibration step that chooses a group-wise threshold. The main theoretical claims are: a high-probability fairness guarantee for the post-processed classifier (Theorem 3), excess-risk bounds of the form |R(bf) - R(f*)| <= dE + |tau*| O(sqrt(log(1/eta)/n)) with dE controlled by squared estimation errors (Theorem 5, Corollary 6), and, as a byproduct, finite-sample excess-risk control for ordinary FLDA with unknown eigenfunctions. Extensive appendices contain the proofs, and the paper reports simulations and an NHANES application with code.

Significance. If the results are correct, this is a timely contribution: it is the first systematic treatment of fairness constraints in functional classification, it addresses the infinite-dimensional obstacles through RKHS/Radon–Nikodym techniques, and it provides explicit finite-sample rates for both fairness and excess risk, including a new FLDA excess-risk result with unknown eigenfunctions. The population characterization is elegant, the post-processing algorithm is simple and reproducible, the proofs are unusually detailed, and the experiments cover Gaussian and non-Gaussian data as well as the perfect-classification regime. The main caveats are proof gaps and quantifier imprecision in the finite-sample theorems rather than a fundamental flaw in the approach.

major comments (3)
  1. [Section 3, Theorem 5] The probability statement "with probability at least 1-eta, eta in (0, eps_pi + eps_eta + eps_tau)" is not a well-defined quantifier as written, because eps_pi, eps_eta and eps_tau are introduced in (8) as bounds that themselves depend on eta through log(1/eta) factors. The admissible range of eta is therefore circular. Please either fix eta in (0,1/2) and specify each eps as an explicit function of eta, or impose an explicit condition such as the one used in Corollary 6 (eta in (0, n^{-1/2} wedge ...)). As stated, the theorem cannot be verified or applied.
  2. [Section 3, Theorem 5, Eq. (9), and Appendix D.1] The step from identity (7) to the bound (9) asserts |D(f*_{D,delta}) - D(bf_{D,delta})| = O(sqrt(log(1/eta)/n)) without proof. This is not an immediate consequence of Theorem 3: the triangle inequality with Theorem 3 gives at most 2 delta + O(sqrt(...)), which is not useful unless delta is small. For DO/PD/DD the claimed bound can be recovered by arguing that, when btau != 0, btau is a first crossing of the monotone step function bD through +/- delta, so that bD(btau) >= delta - O(1/n) (or <= -delta + O(1/n)), and then applying the uniform DKW comparison between bD(btau) and D(bf_{D,delta}). This argument is absent from the appendix. Without it, the advertised bound and the Corollary 6.2 rate |R - R*| <= dE + |tau*| O(N^{-1/2}) are not justified by the written proof.
  3. [Theorems 3 and 5, Definition 4 and Lemma 17] The theorems are stated for any bilinear disparity measure in Definition 4, but the calibration step is only well defined if the set {tau : |bD(tau)| <= delta} is nonempty, and the crossing argument needed for Theorem 5 relies on monotonicity of the empirical function bD. Lemma 17 proves monotonicity only for DO; for PD and DD monotonicity is easy to check but is not stated, and for arbitrary (s_D,a, b_D,a) it is not established. Please either restrict the theorems to the three disparity measures in Definition 3 (with a remark that Theorem 2 gives the population characterization for the full bilinear class), or prove the required empirical monotonicity and range properties for the full class.
minor comments (3)
  1. [Appendix C, proof of Theorem 3] The proof should explicitly record that, conditional on the training data eD, E[bD(tau) | eD] = D(bg_{D,tau}), i.e. the conditional expectation of the empirical disparity is the population disparity of the estimated classifier bg. This identity is the reason the fairness guarantee does not inherit a density-ratio estimation-bias term, and stating it would prevent the reader from misreading the DKW step as insufficient.
  2. [Section 3, Theorem 5, condition (i)] The notation "D(0) /∈ (delta - eps_D, delta ] union [-delta, -delta + eps_D)" should be cleaned up: the interval endpoints and the intended open/closed boundaries should be stated consistently, and the unusual comma in "(delta - eps_D, delta ]" should be removed.
  3. [Section 1.1 and Corollary 6 discussion] The phrase "first time seen" appears several times for the FLDA excess-risk byproduct. The comparison with Wang et al. (2021) is clear, but the claim should be phrased more cautiously unless the authors have verified all antecedent literature beyond the cited works.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the derivation is self-contained, with population thresholds and independent calibration; the flagged proof gap is a correctness issue, not a circular step.

full rationale

I walked the claimed derivation chain: the oracle fair classifier in Theorem 2 is obtained from the generalized Neyman-Pearson lemma applied to the linearized misclassification risk and a bilinear disparity measure; the threshold tau* is a population quantity defined by the optimization in (5), not a fitted value later renamed as a prediction. The plug-in Algorithm 1 estimates (pi, mu, K, eigencomponents) from a separate training set and calibrates btau on an independent calibration set to control the empirical disparity; Theorem 3's population fairness claim is a calibration concentration statement, and Theorem 5/Corollary 6 decompose the excess risk into estimation errors (epsilon_pi, epsilon_eta, epsilon_D) plus a fairness cost. No equation in the paper defines an output as its own input: epsilon_D is an assumed high-probability bound on sup|bD - D|, later instantiated by DKW plus eta-estimation; epsilon_eta is bounded by Lemmas 11-12 via variance-bias decompositions; tau* is not estimated from the quantity Theorem 5 predicts. The only self-citations (Xue et al. 2024; Lin et al. 2023) are contextual or data-descriptive and are not load-bearing. The skeptic's concern, that the Appendix C proof of Theorem 3 bounds DKW fluctuations around E[bD | eD] but does not in the displayed argument bound E[bD | eD] - D, is a potential proof gap or correctness risk rather than a circularity: it concerns an omitted bias term, not a parameter fitted to the target quantity. Therefore no circular step is exhibited, and the score is 0.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The central results use standard FDA regularity assumptions (Assumptions 1 and 2) and classical tools (generalized Neyman-Pearson lemma, DKW inequality, sub-Gaussian concentration, Hall-Horowitz eigenfunction estimation). No free model parameters are fitted in the theory; α and β are unknown structural constants, and the truncation level J is an input subject to constraints. No new particles, forces, or entities are introduced.

assumptions (4)
  • standard math Generalized Neyman-Pearson lemma (Lemma 42) characterizes Bayes-optimal classifiers under linear constraints.
    Used in Theorem 2 to derive the fair Bayes classifier; the lemma is stated and proved in Appendix G.
  • standard math For Gaussian processes, Pa,1 and Pa,0 are mutually absolutely continuous iff µa,1 - µa,0 lies in the RKHS H(Ka); the Radon-Nikodym derivative has the given exponential form.
    Invoked in Section 2.3 to evaluate ηa(X); standard result from Berrendero et al. (2018).
  • domain assumption Assumptions 1 and 2: class probabilities bounded away from 0/1; eigenvalues decay as j^{-α}, α>1, with spectral gap; mean difference coefficients decay as j^{-β}, β>(α+1)/2; SNR lower bound.
    These structural conditions control eigenfunction estimation and ensure the RN derivative has finite tail sum; they define the regime of analysis.
  • standard math Dvoretzky-Kiefer-Wolfowitz inequality and sub-Gaussian/sub-exponential concentration bounds.
    Used extensively in Appendices C-F to control empirical processes and estimation errors.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Fairness-aware Bayes optimal functional classification." pith.science (2026). https://pith.science/paper/USV2J3YH

@misc{pith2026250509471,
  author       = {Pith},
  title        = {Pith review of: Fairness-aware Bayes optimal functional classification},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/USV2J3YH}},
  note         = {Machine review of arXiv:2505.09471}
}
read the original abstract

Algorithmic fairness has become a central topic in machine learning, and mitigating disparities across different subpopulations has emerged as a rapidly growing research area. In this paper, we systematically study the classification of functional data under fairness constraints, ensuring the disparity level of the classifier is controlled below a pre-specified threshold. We propose a unified framework for fairness-aware functional classification, tackling an infinite-dimensional functional space, addressing key challenges from the absence of density ratios and intractability of posterior probabilities, and discussing unique phenomena in functional classification. We further design a post-processing algorithm, Fair Functional Linear Discriminant Analysis classifier (Fair-FLDA), which targets at homoscedastic Gaussian processes and achieves fairness via group-wise thresholding. Under weak structural assumptions on eigenspace, theoretical guarantees on fairness and excess risk controls are established. As a byproduct, our results cover the excess risk control of the standard FLDA as a special case, which, to the best of our knowledge, is first time seen. Our theoretical findings are complemented by extensive numerical experiments on synthetic and real datasets, highlighting the practicality of our designed algorithm.

Figures

Figures reproduced from arXiv: 2505.09471 by the authors.

Figure 1
Figure 1. Effects of steepness of disparity levels on the estimation error of [PITH_FULL_IMAGE:figures/full_fig_p010_1.png] view at source ↗
Figure 2
Figure 2. All x-axis are the values of δ. From left to right: medians of classification errors, medians and 95% quantiles of the disparity measures, in the simulations (1st-3rd columns) and real data (4-6th columns). From top to bottom: DD, DO and PD. Orange dots: FLDA; blue stars: Fair-FLDA; pink triangles: Fair-FLDAc; red solid line: oracle Bayes classifier; grey dashed line: y = x. competitive classification accuracy. 5 Co… view at source ↗
Figure 3
Figure 3. Disparity DO results under the Gaussian model, [PITH_FULL_IMAGE:figures/full_fig_p018_3.png] view at source ↗
Figures from the paper (17 more)
Figure 4
Figure 4. Figure 4: Disparity PD results under the Gaussian model, [PITH_FULL_IMAGE:figures/full_fig_p019_4.png]
Figure 5
Figure 5. Figure 5: Disparity DD results under the Gaussian model, [PITH_FULL_IMAGE:figures/full_fig_p020_5.png]
Figure 6
Figure 6. Figure 6: Error-unfairness trade-off for DO under the Gaussian model, [PITH_FULL_IMAGE:figures/full_fig_p020_6.png]
Figure 7
Figure 7. Figure 7: Error-unfairness trade-off for PD under the Gaussian model, [PITH_FULL_IMAGE:figures/full_fig_p021_7.png]
Figure 8
Figure 8. Figure 8: Error-unfairness trade-off for DD under the Gaussian model, [PITH_FULL_IMAGE:figures/full_fig_p021_8.png]
Figure 9
Figure 9. Figure 9: Disparity DO results under the Gaussian model, [PITH_FULL_IMAGE:figures/full_fig_p022_9.png]
Figure 10
Figure 10. Figure 10: Disparity PD results under the Gaussian model, [PITH_FULL_IMAGE:figures/full_fig_p022_10.png]
Figure 11
Figure 11. Figure 11: Disparity DD results under the Gaussian model, [PITH_FULL_IMAGE:figures/full_fig_p023_11.png]
Figure 12
Figure 12. Figure 12: Disparity DO results under the non-Gaussian model, [PITH_FULL_IMAGE:figures/full_fig_p023_12.png]
Figure 13
Figure 13. Figure 13: Disparity PD results under the non-Gaussian model, [PITH_FULL_IMAGE:figures/full_fig_p024_13.png]
Figure 14
Figure 14. Figure 14: Disparity DD results under the non-Gaussian model, [PITH_FULL_IMAGE:figures/full_fig_p024_14.png]
Figure 15
Figure 15. Figure 15: Oracle disparity DD versus τ . Left: (I); right: (II). A.4 Additional results for real data In practice, we recommend tuning the calibration parameter κ in Fair-FLDAc to achieve more reliable probabilistic disparity control. Specifically, we select the smallest value …
Figure 16
Figure 16. Figure 16: Disparity DO results under (I). Top: n = 1000; middle: n = 2000; bottom: n = 5000. 0.000 0.002 0.004 0.0 0.2 0.4 δ Error 0.00 0.01 0.02 0.0 0.2 0.4 δ U P D ,50 0.00 0.01 0.02 0.0 0.2 0.4 δ U P D ,95 0.000 0.002 0.004 0.0 0.2 0.4 δ Error 0.00 0.01 0.02 0.0 0.2 0.4 δ U …
Figure 17
Figure 17. Figure 17: Disparity PD results under (I). Top: n = 1000; middle: n = 2000; bottom: n = 5000. 26 [PITH_FULL_IMAGE:figures/full_fig_p026_17.png]
Figure 18
Figure 18. Figure 18: Disparity DD results under (I). Top: n = 1000; middle: n = 2000; bottom: n = 5000. 0.000 0.005 0.010 0.0 0.2 0.4 δ Error 0.00 0.05 0.10 0.0 0.2 0.4 δ U D D ,50 0.00 0.05 0.10 0.0 0.2 0.4 δ U D D ,95 0.000 0.005 0.010 0.0 0.2 0.4 δ Error 0.00 0.05 0.10 0.0 0.2 0.4 δ U …
Figure 19
Figure 19. Figure 19: Disparity DD results under (II). Top: n = 1000; middle: n = 2000; bottom: n = 5000. 27 [PITH_FULL_IMAGE:figures/full_fig_p027_19.png]
Figure 20
Figure 20. Figure 20: Results under NHANES with tuned calibration parameters over 100 Monte Carlo trials. [PITH_FULL_IMAGE:figures/full_fig_p028_20.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

48 extracted references · 37 canonical work pages

  1. [1]

    Angwin, J., Larson, J., Mattu, S., and Kirchner, L. (2022). Machine bias. In Ethics of data and analytics , pages 254--264. Auerbach Publications

    work page 2022

  2. [2]

    R., Cuevas, A., and Torrecilla, J

    Berrendero, J. R., Cuevas, A., and Torrecilla, J. L. (2018). On the use of reproducing kernel hilbert spaces in functional classification. Journal of the American Statistical Association , 113(523):1210--1218

    work page 2018

  3. [3]

    Bongiorno, E. G. and Goia, A. (2016). Classification methods for hilbert data based on surrogate density. Computational statistics & data analysis , 99:204--222

    work page 2016

  4. [4]

    Bosq, D. (2000). Linear processes in function spaces: theory and applications , volume 149. Springer Science & Business Media

    2000

  5. [5]

    Cai, T. T. and Zhang, L. (2019). High dimensional linear discriminant analysis: optimality, adaptive algorithm and missing data. Journal of the Royal Statistical Society Series B: Statistical Methodology , 81(4):675--705

    work page 2019

  6. [6]

    Calmon, F., Wei, D., Vinzamuri, B., Natesan Ramamurthy, K., and Varshney, K. R. (2017). Optimized pre-processing for discrimination prevention. Advances in neural information processing systems , 30

    work page 2017

  7. [7]

    National health and nutrition examination survey

    CDC (2006). National health and nutrition examination survey. https://wwwn.cdc.gov/nchs/nhanes/ContinuousNhanes/Default.aspx?BeginYear=2005

    work page 2006

  8. [8]

    E., Huang, L., Keswani, V., and Vishnoi, N

    Celis, L. E., Huang, L., Keswani, V., and Vishnoi, N. K. (2019). Classification with fairness constraints: A meta-algorithm with provable guarantees. In Proceedings of the conference on fairness, accountability, and transparency , pages 319--328

    work page 2019

Show all 48 references
  1. [9]

    Cho, J., Hwang, G., and Suh, C. (2020). A fair classifier using kernel density estimation. Advances in neural information processing systems , 33:15088--15099

    2020

  2. [10]

    and Schreuder, N

    Chzhen, E. and Schreuder, N. (2022). A minimax framework for quantifying risk-fairness trade-off in regression. The Annals of Statistics , 50(4):2416--2442

    2022

  3. [11]

    Coffey, N., Hinde, J., and Holian, E. (2014). Clustering longitudinal profiles using p-splines and mixed effects models applied to time-course gene expression data. Computational Statistics & Data Analysis , 71:14--29

    2014

  4. [12]

    Corbett-Davies, S., Pierson, E., Feller, A., Goel, S., and Huq, A. (2017). Algorithmic decision making and the cost of fairness. In Proceedings of the 23rd acm sigkdd international conference on knowledge discovery and data mining , pages 797--806

    2017

  5. [13]

    Dai, X., M \"u ller, H.-G., and Yao, F. (2017). Optimal bayes classifiers for functional data and density ratios. Biometrika , 104(3):545--560

    2017

  6. [14]

    and Hall, P

    Delaigle, A. and Hall, P. (2012). Achieving near perfect classification for functional data. Journal of the Royal Statistical Society Series B: Statistical Methodology , 74(2):267--286

    2012

  7. [15]

    W., Pollard, D., and Zhou, H

    Dou, W. W., Pollard, D., and Zhou, H. H. (2012). Estimation in functional regression for general exponential families. The Annals of Statistics , pages 2421--2451

    2012

  8. [16]

    Dvoretzky, A., Kiefer, J., and Wolfowitz, J. (1956). Asymptotic minimax character of the sample distribution function and of the classical multinomial estimator. The Annals of Mathematical Statistics , pages 642--669

    1956

  9. [17]

    Big data: A report on algorithmic systems, opportunity, and civil rights

    Executive Office of the President (2016). Big data: A report on algorithmic systems, opportunity, and civil rights. https://obamawhitehouse.archives.gov/sites/default/files/microsites/ostp/2016_0504_data_discrimination.pdf

    2016

  10. [18]

    and Sakuma, J

    Fukuchi, K. and Sakuma, J. (2023). Demographic parity constrained minimax optimal regression under linear model. Advances in Neural Information Processing Systems , 36:8653--8689

    2023

  11. [19]

    and Horowitz, J

    Hall, P. and Horowitz, J. L. (2007). Methodology and convergence rates for functional linear regression . The Annals of Statistics , 35(1):70 -- 91

    2007

  12. [20]

    Hardt, M., Price, E., and Srebro, N. (2016). Equality of opportunity in supervised learning. Advances in neural information processing systems , 29

    2016

  13. [21]

    Heinrichs, F., Heim, M., and Weber, C. (2023). Functional neural networks: Shift invariant models for functional data with applications to eeg classification. In International Conference on Machine Learning , pages 12866--12881. PMLR

    2023

  14. [22]

    and Zhang, L

    Hou, X. and Zhang, L. (2024). Finite-sample and distribution-free fair classification: Optimal trade-off between excess risk and fairness, and the cost of group-blindness. arXiv preprint arXiv:2410.16477

    2024 arXiv

  15. [23]

    Hubert, M., Rousseeuw, P., and Segaert, P. (2017). Multivariate and functional classification using depth and distance. Advances in Data Analysis and Classification , 11:445--466

    2017

  16. [24]

    Jiang, R., Pacchiano, A., Stepleton, T., Jiang, H., and Chiappa, S. (2020). Wasserstein fair classification. In Uncertainty in artificial intelligence , pages 862--872. PMLR

    2020

  17. [25]

    Johndrow, J. E. and Lum, K. (2019). An algorithm for removing sensitive information. The Annals of Applied Statistics , 13(1):189--220

    2019

  18. [26]

    P., Ghorbani, A., and Zou, J

    Kim, M. P., Ghorbani, A., and Zou, J. (2019). Multiaccuracy: Black-box post-processing for fairness in classification. In Proceedings of the 2019 AAAI/ACM Conference on AI, Ethics, and Society , pages 247--254

    2019

  19. [27]

    and Stefanucci, M

    Kraus, D. and Stefanucci, M. (2019). Classification of functional fragments by regularized linear classifiers with domain selection. Biometrika , 106(1):161--180

    2019

  20. [28]

    Li, P., Zou, J., and Zhang, L. (2022). FaiREE : Fair classification with finite-sample and distribution-free guarantee. arXiv preprint arXiv:2211.15072

    2022 arXiv

  21. [29]

    Lila, E., Zhang, W., and Rane Levendovszky, S. (2024). Interpretable discriminant analysis for functional data supported on random nonlinear domains with an application to alzheimer’s disease. Journal of the Royal Statistical Society Series B: Statistical Methodology , 86(4):1...

    2024

  22. [30]

    Lin, Z., Kong, D., and Wang, L. (2023). Causal inference on distribution functions. Journal of the Royal Statistical Society Series B: Statistical Methodology , 85(2):378--398

    2023

  23. [31]

    Massart, P. (1990). The tight constant in the dvoretzky-kiefer-wolfowitz inequality. The annals of Probability , pages 1269--1283

    1990

  24. [32]

    Meister, A. (2016). Optimal classification and nonparametric regression for functional data . Bernoulli , 22(3):1729 -- 1744

    2016

  25. [33]

    Narasimhan, H. (2018). Learning with complex loss functions and constraints. In International Conference on Artificial Intelligence and Statistics , pages 1646--1654. PMLR

    2018

  26. [34]

    Preda, C., Saporta, G., and L \'e v \'e der, C. (2007). Pls classification of functional data. Computational Statistics , 22(2):223--235

    2007

  27. [35]

    Qiao, X., Guo, S., and James, G. M. (2019). Functional graphical models. Journal of the American Statistical Association , 114(525):211--222

    2019

  28. [36]

    L., Ramos-Carre \ n o, C., Sanchez-Montanes, M., and Su \'a rez, A

    Torrecilla, J. L., Ramos-Carre \ n o, C., Sanchez-Montanes, M., and Su \'a rez, A. (2020). Optimal classification of gaussian processes in homo-and heteroscedastic settings. Statistics and Computing , 30(4):1091--1111

    2020

  29. [37]

    Vershynin, R. (2018). High-dimensional probability: An introduction with applications in data science , volume 47. Cambridge university press

    2018

  30. [38]

    Wang, J.-L., Chiou, J.-M., and M \"u ller, H.-G. (2016). Functional data analysis. Annual Review of Statistics and its application , 3(1):257--295

    2016

  31. [39]

    Wang, S., Shang, Z., Cao, G., and Liu, J. (2021). Optimal classification for functional data. arXiv preprint arXiv:2103.00569

    2021 arXiv

  32. [40]

    N., and Calmon, F

    Wei, D., Ramamurthy, K. N., and Calmon, F. P. (2021). Optimized score transformation for consistent fair classification. Journal of Machine Learning Research , 22(258):1--78

    2021

  33. [41]

    C., Li, Z., and Tewari, A

    Wong, K. C., Li, Z., and Tewari, A. (2020). Lasso guarantees for -mixing heavy-tailed time series. Annals of Statistics , 48(2):1124--1142

    2020

  34. [42]

    and Strohmer, T

    Xu, S. and Strohmer, T. (2023). Fair data representation for machine learning at the pareto frontier. Journal of Machine Learning Research , 24(331):1--63

    2023

  35. [43]

    Xue, G., Lin, Z., and Yu, Y. (2024). Optimal estimation in private distributed functional data analysis. arXiv preprint arXiv:2412.06582

    2024

  36. [44]

    Yang, F., Cisse, M., and Koyejo, S. (2020). Fairness with overlapping groups; a probabilistic perspective. Advances in neural information processing systems , 33:4067--4078

    2020

  37. [45]

    Zapata, J., Oh, S.-Y., and Petersen, A. (2022). Partial separability and functional graphical models for multivariate gaussian processes. Biometrika , 109(3):665--681

    2022

  38. [46]

    Zeng, X., Cheng, G., and Dobriban, E. (2024a). Bayes-optimal fair classification with linear disparity constraints via pre-, in-, and post-processing. arXiv preprint arXiv:2402.02817

    2024 arXiv

  39. [47]

    Zeng, X., Cheng, G., and Dobriban, E. (2024b). Minimax optimal fair classification with bounded demographic disparity. arXiv preprint arXiv:2403.18216

    2024 arXiv

  40. [48]

    and Wang, J.-L

    Zhang, X. and Wang, J.-L. (2016). From sparse to dense functional data and beyond . The Annals of Statistics , 44(5):2281--2321

    2016

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.