Noise-Adaptive Conformal Classification with Marginal Coverage

Alessandra Menafoglio; Matteo Sesia; Simone Vantini; Teresa Bortolotti; Xin Tong; Y. X. Rachel Wang

REVIEW 4 major objections 3 minor 25 references

Noise-Adaptive Conformal Classification with Marginal Coverage

T0 review · 4 major / 3 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read This paper introduces an adaptive conformal prediction method that provably maintains marginal coverage when calibration labels are corrupted by random noise, and shows how to compute a finite-sample correction factor that keeps…

desk verdict The marginal-coverage method is a real step forward, but the finite-sample guarantee in Theorem 3 rests on a sign error that invalidates the proof of the headline claim. read the letter →

arxiv 2501.18060 v1 pith:YETHNLEJ submitted 2025-01-29 stat.ME cs.LGstat.ML

Teresa Bortolotti , Y. X. Rachel Wang , Xin Tong , Alessandra Menafoglio , Simone Vantini , Matteo Sesia This is my paper · ORCID

classification stat.MEcs.LGstat.ML MSC 62G1562G2062H30

keywords conformalinferenceclassificationlabelnoisemarginalcoveragepredictionsetsempiricalprocessnon-exchangeabilityadaptivecalibration

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

The paper tackles a practical failure mode of conformal prediction: when the labels used for calibration are contaminated by random noise, standard split-conformal sets become over-conservative, wasting informativeness. The authors propose an adaptive calibration algorithm that estimates the gap between coverage computed with noisy labels and coverage with clean labels, then shifts the prediction threshold by a correction factor. They prove that with the right correction factor the method achieves the target marginal coverage in finite samples, and they supply two computable versions of that factor: a finite-sample upper bound and an asymptotic approximation based on a Gaussian limit process. If correct, this gives practitioners a way to keep 90% (or any target) coverage under label noise without needing clean labels at calibration time.

What carries the argument

The core object is the marginal coverage inflation factor $\Delta$(t) = F(t) - F(tilde-F)(t), and its empirical estimator hat{$\Delta$}(t) built from the inverse transition matrix W = $T^{{-1}}$, the empirical label frequencies, and conditional empirical CDFs of scores. The ideal correction factor delta*(n) is the expected supremum of the empirical process psi-hat(t) = sum_{k,l} W_{kl}(hat{rho}_l hat{F}_l^k(t) - tilde{rho}_l tilde{F}_l^k(t)). The paper bounds delta*(n) in two ways: a finite-sample chaining bound delta_FS(n) using Massart's lemma and Dudley's integral, and an asymptotic approximation delta_asy(n) via a Donsker theorem showing sqrt{n} psi-hat converges to a Generalized Brownian Bridge. These correction factors are what make the adaptive threshold in Algorithm 1 work.

What would settle it

Simulate data with two classes under a two-level randomized response model with known T, run Algorithm 1 with the finite-sample correction delta_FS(n) for n = 1000, 5000, 20000, and verify empirically that marginal coverage stays at or above 1 - alpha. Then repeat the same simulation with a deliberately biased estimate of T (e.g., misspecified noise rate plus 0.05) and check that coverage drops below 1 - alpha by more than Monte Carlo error as n grows; if it does not drop, the method is more robust than the theory claims, while if it drops, the theory's reliance on exact T is confirmed.

Watch

Extended reading notes

Core claim

The central claim is that the effect of random label noise on marginal conformal coverage is fully captured by a scalar function $\Delta$(t) = F(t) - F(tilde-F)(t), the difference between the CDF of non-conformity scores evaluated at the true label and at the observed noisy label. Under the assumption that the label transition matrix T is known and invertible, this function can be written as a linear combination of estimable quantities, and its empirical estimate can be used to correct the conformal threshold. The paper proves that Algorithm 1, run with any correction factor delta(n) at least as large as the expected supremum of a certain zero-mean empirical process, yields prediction sets with P[Y_{n+1} in hat{C}(X_{n+1})] >= 1 - $\alpha$. It then constructs two concrete such factors: delta_FS(n) from a chaining bound that scales as 1/$\sqrt$(n) independent of the number of classes, and delta_asy(n) from a Generalized Brownian Bridge limit whose expectation is estimated by Monte Carlo. Under additional smoothness and accuracy assumptions, coverage is also bounded above by 1 - $\alpha$ + delta(n) + o(1), so the sets are asymptotically tight.

Load-bearing premise

The finite-sample coverage guarantee requires that the label transition matrix T is known exactly and invertible, so the inverse W used in the correction is correct; if T is estimated, the theoretical guarantee no longer applies and the method rests on empirical evidence of robustness.

Editorial extensions

If this is right

If the central claim holds, practitioners can apply conformal prediction to data sets with noisy labels and obtain marginal coverage at the nominal level with prediction sets that shrink toward the noise-free size as n grows.
The 1/sqrt(n) scaling of the finite-sample correction, independent of the number of classes K, implies the method remains informative in many-class or highly imbalanced problems where label-conditional conformal methods produce impractically large sets.
The asymptotic correction delta_asy(n), although not theoretically guaranteed to be conservative in finite samples, performs well in experiments and offers a computationally feasible alternative for large calibration sets.
The method can be applied with an estimated transition matrix T, as demonstrated on real data, suggesting that clean labels may be needed only to estimate T once, not for every calibration set.
The decomposition of coverage inflation via Delta(t) provides a template for adapting conformal inference to other structured deviations from exchangeability that can be modeled as a known invertible transformation.

Reading between the lines

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

A natural extension the authors leave implicit is to treat T as uncertain and replace the point estimate by a confidence set, which would convert the empirical robustness on real data into a rigorous coverage guarantee under model misspecification.
The same Delta(t) correction strategy could be adapted to regression by defining non-conformity scores on continuous labels and modeling label noise through a known corruption density, though the empirical process analysis would need reworking.
The Generalized Brownian Bridge approximation suggests a direct bootstrap alternative: resample calibration scores to estimate the expected supremum of psi-hat empirically, which could avoid the discretization and extrapolation steps and work for small n.
Because the correction is marginal rather than label-conditional, the method implicitly trades per-class guarantees for overall efficiency; users with rare but safety-critical classes should combine it with a separate check on those classes.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

The paper gives the first adaptive conformal method for marginal coverage under label noise that promises a 1/sqrt(n) inflation factor independent of K. The empirical process machinery (chaining, Donsker) is a genuine improvement over the union-bound approach in Sesia et al. 2024, and the experiments on CIFAR-10H and BigEarthNet are thorough, with code available. That part is worth taking seriously.

The problem is the proof of Theorem 3. The authors parametrize the inverse transition matrix W for the randomized response model as beta0 I + (beta_k/K) J with beta0 = -1/(1-eps), beta_k = eps/(1-eps). But the true W is 1/(1-eps) I - eps/(K(1-eps)) J. Their beta gives -W, not W. So the sum beta0 + (sum beta_k)/K equals -1, not 1, and the step E[sup psi1] = beta0 c(n) is false when beta0 is negative. The same decomposition drives the finite-sample bound delta_FS in Theorem 4, so that result is also unsupported. In other words, the advertised guarantee that delta_FS(n) >= delta*(n) does not go through as written.

I checked the proof carefully and the stress-test note is right: this is not a stylistic complaint, it's a load-bearing sign error. The asymptotic method in Section 4 uses W directly and is not affected, and the empirical results look good, but the paper's central theoretical contribution—the finite-sample coverage guarantee—is unproven in this version. The authors also honestly acknowledge that the real-data experiments use estimated T rather than known T, which is a lesser but real caveat.

Who should read this: anyone working on conformal prediction with noisy labels. The method is likely salvageable by flipping the signs in the parametrization and re-running the constants, and the empirical evidence is encouraging. So I would send it to a serious referee, but with the expectation of major revision: the proof needs to be corrected before the guarantee can be trusted. My own verdict would be reject in current form, not because the idea is bad, but because the main theorem is not proved.

Referee Report

4 major / 3 minor

Summary. The paper proposes a split-conformal classification method that aims to maintain marginal coverage when calibration labels are corrupted by random label noise. Under Assumptions 1–2 (noise conditionally independent of features given the true label, and a known invertible transition matrix T), the method estimates the coverage inflation factor Δ(t) and chooses a threshold using a correction factor δ(n); Theorem 2 states that coverage ≥ 1 − α whenever δ(n) ≥ δ*(n). Sections 3–4 propose two ways of choosing δ(n): a finite-sample upper bound δFS(n) based on the constant c(n) plus a chaining error term, and an asymptotic Monte Carlo approximation. Section 5 gives an upper coverage bound under additional assumptions. The method is evaluated on synthetic data and on CIFAR-10H and BigEarthNet. The advertised central contribution is a finite-sample marginal coverage guarantee with a correction factor that decays as 1/√n independently of the number of classes.

Significance. If the main theorems were correct, the paper would offer a useful and nontrivial extension of conformal inference under label noise, and the comparison to label-conditional methods would be informative. Credit is due for the clean structural result in Theorem 2, the Donsker-style asymptotic argument in Theorem 5, and the public code and extensive experiments. However, the central theoretical result is not established: the proof of Theorem 3 contains a sign error, and the same error undermines Theorem 4, so the proposed finite-sample correction δFS(n) is not proven to be conservative. Since the advertised coverage guarantee depends on δ(n) ≥ δ*(n), the paper's main claim is unsupported. The empirical results do not repair the proof.

major comments (4)

[Appendix C.3 / Section 3.1] In the proof of Theorem 3, the randomized-response inverse is W = 1/(1−ε)I − ε/[K(1−ε)]J, but the proof sets β0 = −1/(1−ε) and β_k = ε/(1−ε), so β0 I + (β_k/K)J equals −W, not W. Consequently β0 + (1/K)Σ_k β_k = −1, contradicting the proof's claim that this sum equals 1. The subsequent identities E sup ψ1 = β0 c(n) and E sup ψ2 ≤ (Σ β_k/K)c(n) require nonnegative coefficients, whereas here β0 < 0. Hence the conclusion c(n) ≥ δ*(n) is not proven.
[Section 3.2, Eq. (14), and Appendix C.3] The same sign issue invalidates the proof of Theorem 4. The first term in δFS(n) is c(n)(β0 + Σ_k β_k/K) without absolute values, and the derivation uses E sup ψ1 = β0 c(n) and E sup ψ2 ≤ (Σ_k β_k/K)c(n), which are only valid for nonnegative β0 and β_k. The minimization over all β ∈ R^{K+1} can exploit negative values of the first term, so the argument does not establish δFS(n) ≥ δ*(n). Additionally, Assumption 3 is stated in Theorem 4 but is not used in the proof.
[Section 5 (Assumption 6) and Appendix C.5] There is a formal inconsistency in the definition of d(n): Assumption 6 in the main text defines d(n) = n^{−1/4} · inf_β {√(π/2)(|β0| + Σ_k|β_k|/K) + B(K,n,β)}, while the proof of Theorem 6 defines d(n) = inf_β {1/(4√n)√(π/2)(|β0| + Σ_k|β_k|/K) + 1/(4√n)B(K,n,β)}. Because the upper bound in Theorem 6 depends on d(n), the assumption and the proof are not aligned.
[Section 7] In the real-data experiments, the transition matrix T is estimated from training data, so Assumption 2 fails and the finite-sample coverage guarantee of Theorem 2 does not apply. The paper acknowledges this limitation, but as a result the empirical coverage plots, while encouraging, are not direct evidence for the main theoretical claim.

minor comments (3)

[Appendix C.1 (Proof of Theorem 1)] The notation for the clean and noisy label frequencies is mistyped: both are written as ~ρ_k; the clean-label frequency should be denoted ρ_k and the noisy-label frequency ~ρ_k.
[Theorem 4 / Assumption 3] Assumption 3 is stated as a hypothesis but is never invoked in the proof of Theorem 4 or in Lemma A1; the authors should either use it or remove it.
[Section 4 / Appendix B.5] The rule-of-thumb choices for the Monte Carlo parameters h, M, and the Richardson extrapolation order are presented only in the appendix; stating them in the main text would make the asymptotic implementation reproducible without reading the supplement.

Circularity Check

0 steps flagged · score 1.0 of 10

No material circularity: the adaptive correction factors are derived as distribution-free upper bounds on an empirical-process supremum, not fitted to the target coverage; the flagged Theorem 3 sign error is a correctness gap, not a circular reduction.

full rationale

The central derivation chain is self-contained. The estimator Δ̂(t) is an empirical counterpart of the coverage inflation factor Δ(t) = F(t) − F̃(t), and Algorithm 1's threshold condition uses δ(n) only as an additive correction. Theorem 2 proves coverage when δ(n) ≥ δ*(n), where δ*(n) is the expected supremum of the empirical process in (13); that quantity does not depend on α and is not fitted to force coverage. Section 3 derives δFS(n) as an upper bound on δ*(n) using Massart's lemma, Dudley's theorem, and chaining, while Section 4 obtains δ̂asy from Donsker theory; neither correction is calibrated to the observed coverage outcome. Citations to Sesia et al. (2024) are for notation, background, and baseline comparisons, and they are not load-bearing for Theorems 2–4. The substantive issue in the derivation chain is a sign error in the proof of Theorem 3: Appendix C.3 defines β0 := −1/(1−ϵ) and β_k := ϵ/(1−ϵ), which makes β0 I + β_k/K J the negative of W and contradicts the claimed β0 + (1/K)Σβ_k = 1; consequently the step E sup ψ1 = β0 c(n) is invalid for negative β0. That is a correctness and mathematical-support gap, not circularity, because the claimed bound is not an input to its own derivation. The paper also validates the method against simulations and external real data with available code, so the appropriate circularity score is 1.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central validity theorem depends on the label-noise model (T known, \tilde Y independent of X given Y) and on the distribution-free bound for the empirical process. The finite-sample bound introduces the β parametrization and smoothness assumption. The upper-bound tightness theorem adds three more technical assumptions.

free parameters (3)

correction factor δ(n) = δFS(n) or \hat δ_asy(n)
User-chosen; coverage holds only if δ(n) ≥ δ*(n). The finite-sample bound is an upper bound; the asymptotic version is an approximation without a finite-sample guarantee.
β coefficients in decomposition of W (K+1 values) = optimized via convex program (Appendix B.3)
Introduced to split the empirical process into a well-behaved part and a remainder; the finite-sample correction δFS(n) is the minimum over β.
Monte Carlo grid h and sample size M (asymptotic correction) = rule of thumb in Appendix B.5
Affects the accuracy of the estimated expected supremum of the Gaussian process; not fitted to coverage.

assumptions (4)

domain assumption \tilde Y is conditionally independent of X given Y (Assumption 1)
Needed for expression (7) of Δ(t) in terms of estimable quantities; the paper tests robustness empirically when it is violated.
domain assumption The label transition matrix T is known and invertible (Assumption 2)
Algorithm 1 takes W = T^{-1} as input; theory does not cover estimated T.
domain assumption Score CDFs have bounded densities with positive infimum (Assumption 3)
Used in the covering-number and chaining arguments in Lemma A1 and Theorem 4; can be induced by adding noise to scores.
ad hoc to paper Assumptions 4-6 for the tightness upper bound (Theorem 6)
Assumption 5 is a stochastic dominance condition that the correct label tends to have smaller score; Assumption 6 requires the standard method to be sufficiently conservative. These only affect the upper bound on over-coverage, not the validity guarantee.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Noise-Adaptive Conformal Classification with Marginal Coverage." pith.science (2026). https://pith.science/paper/YETHNLEJ

@misc{pith2026250118060,
  author       = {Pith},
  title        = {Pith review of: Noise-Adaptive Conformal Classification with Marginal Coverage},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YETHNLEJ}},
  note         = {Machine review of arXiv:2501.18060}
}

read the original abstract

Conformal inference provides a rigorous statistical framework for uncertainty quantification in machine learning, enabling well-calibrated prediction sets with precise coverage guarantees for any classification model. However, its reliance on the idealized assumption of perfect data exchangeability limits its effectiveness in the presence of real-world complications, such as low-quality labels -- a widespread issue in modern large-scale data sets. This work tackles this open problem by introducing an adaptive conformal inference method capable of efficiently handling deviations from exchangeability caused by random label noise, leading to informative prediction sets with tight marginal coverage guarantees even in those challenging scenarios. We validate our method through extensive numerical experiments demonstrating its effectiveness on synthetic and real data sets, including CIFAR-10H and BigEarthNet.

Figures

Figures reproduced from arXiv: 2501.18060 by the authors.

**Figure 2.** Performances of different conformal methods on simulated data with labels contaminated [PITH_FULL_IMAGE:figures/full_fig_p015_2.png] view at source ↗

**Figure 3.** Performances of conformal prediction methods on simulated data with contaminated [PITH_FULL_IMAGE:figures/full_fig_p016_3.png] view at source ↗

Figures from the paper (3 more)

**Figure 4.** Figure 4: Performances of conformal prediction methods on simulated data with contaminated [PITH_FULL_IMAGE:figures/full_fig_p017_4.png]

**Figure 5.** Figure 5: Performances of the conformal prediction methods on the CIFAR-10H data set with [PITH_FULL_IMAGE:figures/full_fig_p018_5.png]

**Figure 6.** Figure 6: Performance of conformal prediction methods on the BigEarthNet data set with contam [PITH_FULL_IMAGE:figures/full_fig_p020_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

25 extracted references · 22 canonical work pages

[1]

write newline

" write newline "" before.all 'output.state := FUNCTION article output.bibitem format.authors "author" output.check author format.key output output.year.check new.block format.title "title" output.check new.block crossref missing format.jour.vol output format.article.crossref output.nonnull format.pages output if new.block note output fin.entry FUNCTION b...
[2]

Barber, R. F., E. J. Cand \`e s, A. Ramdas, and R. J. Tibshirani (2023). Conformal prediction beyond exchangeability . Ass. Stat.\/ 51\/ (2), 816--845

work page 2023
[3]

Clarkson, J., W. Xu, M. Cucuringu, and G. Reinert (2024). Split conformal prediction under data contamination. In Proc. 13th Symp. Conformal Probabilistic Prediction Appl. , Volume 230 of Proceedings of Machine Learning Research , pp.\ 5--27. PMLR

work page 2024
[4]

Clasen, K. N., L. Hackel, T. Burgert, G. Sumbul, B. Demir, and V. Markl (2024). REBEN : Refined BigEarthNet dataset for remote sensing image analysis. arXiv preprint arXiv:2407.03653\/

arXiv 2024
[5]

Angelopoulos, S

Ding, T., A. Angelopoulos, S. Bates, M. Jordan, and R. J. Tibshirani (2024). Class-conditional conformal prediction with many classes. Adv. Neural Inf. Process. Syst.\/ 36

work page 2024
[6]

McSherry, K

Dwork, C., F. McSherry, K. Nissim, and A. Smith (2006). Calibrating noise to sensitivity in private data analysis. In Theory of Cryptography , pp.\ 265--284. Springer

work page 2006
[7]

Feldman, S

Einbinder, B.-S., S. Feldman, S. Bates, A. N. Angelopoulos, A. Gendler, and Y. Romano (2024). Label Noise Robustness of Conformal Prediction . J. Mach. Learn. Res.\/ 25\/ (328), 1--66

work page 2024
[8]

Brack, and T

Ko mann, D., V. Brack, and T. Wilhelm (2022). Seasonet: A seasonal scene classification, segmentation and retrieval dataset for satellite imagery over germany. IGARSS 2022 - IEEE Int. Geosci. Remote Sens. Symp.\/ , 243--246

work page 2022

Show all 25 references

[9]

Robins, and L

Lei, J., J. Robins, and L. Wasserman (2013). Distribution-free prediction sets. J. Am . Stat. Assoc.\/ 108\/ (501), 278--287

2013
[10]

Massart, P. (2000). Some applications of concentration inequalities to statistics. In Ann. Fac. Sci. Toulouse Math. , Volume 9, pp.\ 245--303

2000
[11]

Natalini, P. and B. Palumbo (2000). Inequalities for the incomplete gamma function. Mathematical Inequalities & Applications\/ 3\/ (1), 69--67

2000
[12]

Varoquaux, A

Pedregosa, F., G. Varoquaux, A. Gramfort, V. Michel, B. Thirion, O. Grisel, M. Blondel, P. Prettenhofer, R. Weiss, V. Dubourg, et al. (2011). Scikit-learn: Machine learning in Python . J. Mach. Learn. Res.\/ 12 , 2825--2830

2011
[13]

Penso, C. and J. Goldberger (2024). A Conformal Prediction Score that is Robust to Label Noise . arXiv preprint arXiv:2405.02648\/

2024 arXiv
[14]

Peterson, J. C., R. M. Battleday, T. L. Griffiths, and O. Russakovsky (2019). Human uncertainty makes classification more robust. In Proc. IEEE/CVF Int. Conf. Comput. Vis. , pp.\ 9617--9626

2019
[15]

and P.-L

Pinto, J. and P.-L. St-Charles (2022). BigEarthNet . https://github.com/jerpint/bigearthnet

2022
[16]

Richardson, L. F. and R. T. Glazebrook (1911). The approximate arithmetical solution by finite differences of physical problems involving differential equations, with an application to the stresses in a masonry dam . Phil. Trans. R. Soc. A\/ 210\/ (459-470), 307--357

1911
[17]

Sesia, and E

Romano, Y., M. Sesia, and E. Cand \`e s (2020). Classification with valid and adaptive coverage. Adv. in Neural Inf. Process. Syst.\/ 33 , 3581--3591

2020
[18]

Schmitt, M., L. H. Hughes, C. Qiu, and X. X. Zhu (2019). SEN12MS -- A curated dataset of georeferenced multi-spectral sentinel-1/2 imagery for deep learning and data fusion . ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci. \/ IV-2/W7 , 153--160

2019
[19]

Sesia, M., Y. R. Wang, and X. Tong (2024). Adaptive conformal classification with noisy labels . J. R. Stat. Soc. (B)\/ , qkae114

2024
[20]

Charfuelan, B

Sumbul, G., M. Charfuelan, B. Demir, and V. Markl (2019). Bigearthnet: A large-scale benchmark archive for remote sensing image understanding . In IGARSS 2019 - 2019 IEEE International Geoscience and Remote Sensing Symposium , pp.\ 5901--5904. IEEE

2019
[21]

Van der Vaart, A. W. (2000). Asymptotic statistics , Volume 3. Cambridge University Press

2000
[22]

Gammerman, and G

Vovk, V., A. Gammerman, and G. Shafer (2005). Algorithmic Learning in a Random World . Springer Science & Business Media, 2005

2005
[23]

Lindsay, I

Vovk, V., D. Lindsay, I. Nouretdinov, and A. Gammerman (2003). Mondrian confidence machine. Technical Report\/

2003
[24]

Wainwright, M. J. (2019). High-dimensional statistics: A non-asymptotic viewpoint , Volume 48. Cambridge University Press

2019
[25]

Warner, S. L. (1965). Randomized response: A survey technique for eliminating evasive answer bias . J. Am . Stat. Assoc.\/ 60\/ (309), 63--69

1965

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.

[1] [1]

write newline

" write newline "" before.all 'output.state := FUNCTION article output.bibitem format.authors "author" output.check author format.key output output.year.check new.block format.title "title" output.check new.block crossref missing format.jour.vol output format.article.crossref output.nonnull format.pages output if new.block note output fin.entry FUNCTION b...

[2] [2]

Barber, R. F., E. J. Cand \`e s, A. Ramdas, and R. J. Tibshirani (2023). Conformal prediction beyond exchangeability . Ass. Stat.\/ 51\/ (2), 816--845

work page 2023

[3] [3]

Clarkson, J., W. Xu, M. Cucuringu, and G. Reinert (2024). Split conformal prediction under data contamination. In Proc. 13th Symp. Conformal Probabilistic Prediction Appl. , Volume 230 of Proceedings of Machine Learning Research , pp.\ 5--27. PMLR

work page 2024

[4] [4]

Clasen, K. N., L. Hackel, T. Burgert, G. Sumbul, B. Demir, and V. Markl (2024). REBEN : Refined BigEarthNet dataset for remote sensing image analysis. arXiv preprint arXiv:2407.03653\/

arXiv 2024

[5] [5]

Angelopoulos, S

Ding, T., A. Angelopoulos, S. Bates, M. Jordan, and R. J. Tibshirani (2024). Class-conditional conformal prediction with many classes. Adv. Neural Inf. Process. Syst.\/ 36

work page 2024

[6] [6]

McSherry, K

Dwork, C., F. McSherry, K. Nissim, and A. Smith (2006). Calibrating noise to sensitivity in private data analysis. In Theory of Cryptography , pp.\ 265--284. Springer

work page 2006

[7] [7]

Feldman, S

Einbinder, B.-S., S. Feldman, S. Bates, A. N. Angelopoulos, A. Gendler, and Y. Romano (2024). Label Noise Robustness of Conformal Prediction . J. Mach. Learn. Res.\/ 25\/ (328), 1--66

work page 2024

[8] [8]

Brack, and T

Ko mann, D., V. Brack, and T. Wilhelm (2022). Seasonet: A seasonal scene classification, segmentation and retrieval dataset for satellite imagery over germany. IGARSS 2022 - IEEE Int. Geosci. Remote Sens. Symp.\/ , 243--246

work page 2022