Navigating Towards Fairness with Data Selection

Fang Chen; Feng Zhou; Xuhui Fan; Yang Wang; Yixuan Zhang; Zhidong Li

REVIEW 2 major objections 5 minor 45 references

Navigating Towards Fairness with Data Selection

T0 review · 2 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read This paper claims that a data selection rule scoring training points by their expected impact on a fair distribution—using a zero-shot predictor as a proxy for a clean holdout set and a peer prediction penalty—can pick instances less…

desk verdict The fairness-aware data selection idea is new, but Eq. (13) is underived; as written the central claim fails, though the heuristic may be salvageable with a corrected derivation. read the letter →

arxiv 2412.11072 v1 pith:A2PYG2D3 submitted 2024-12-15 cs.LG cs.CY

Yixuan Zhang , Zhidong Li , Yang Wang , Fang Chen , Xuhui Fan , Feng Zhou This is my paper · ORCID

classification cs.LGcs.CY

keywords fairdataselectionlabelbiaszero-shotpredictorpeerpredictiondemographicparityfairnessonlinebatch

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 tries to establish that a data selection rule, which scores each training example by its expected effect on a fair predictive distribution, can train fair classifiers even when the labels themselves are biased against a sensitive group. It claims the score can be computed without a clean holdout set: a zero-shot pretrained model such as CLIP stands in for the validation model, and a peer prediction penalty keeps that proxy from leaking group bias into the scores. The resulting objective is a drop-in filter for any log-likelihood or cross-entropy based classifier, needs no noise-rate estimation, and in experiments on CelebA and LFW+a with 20% and 40% symmetric label bias it improves both test accuracy and demographic parity relative to baselines while converging faster.

What carries the argument

The engine of the method is the selection objective in Eq. (13), which assigns each candidate sample $(x,y,s)$ the score $L[y|x,D_t,s] + (1-\alpha) L[y,\tilde{f}(x),s] - \gamma\,\mathbb{E}_{Y|D_{s'}}[L[Y,\tilde{f}(x),s]]$. Here $L[y|x,D_t,s]$ is the current model's training loss; the zero-shot predictor $\tilde{f}$ (CLIP in the experiments) supplies the holdout-loss term that replaces a clean validation set; and the peer prediction expectation, drawn from a different demographic group $s'$, penalizes the proxy's group-dependent loss. The companion decomposition of the expected score into a fair-model term, a noisy-loss penalty, and a demographic-disagreement penalty is what the authors use to argue that maximizing the score prefers instances less affected by label bias.

What would settle it

On a benchmark with known clean labels, construct a version of the task where CLIP has group-dependent errors (e.g., its predictions for female images are systematically shifted toward the majority label), then run Eq. (13) selection and measure the fraction of selected instances whose true clean labels differ from their observed biased labels. If that fraction is not materially lower than uniform sampling, or if training on the selected set does not reduce demographic parity violation, the central claim fails.

Watch

Extended reading notes

Core claim

The central claim is that the reducible holdout loss (RHO-LOSS) selection criterion can be re-derived with a fair label distribution in place of the observed biased one, and that the resulting selection function is tractable without a holdout set. The ideal selection objective is written as a group-weighted score, and the intractable term involving the holdout posterior is lower-bounded and then approximated by a zero-shot predictor $\tilde{f}$ under the assumption that the predictor's training data make its posterior narrow. Adding a cross-group peer loss yields the final selection score $L[y|x,D_t,s] + (1-\alpha)L[y,\tilde{f}(x),s] - \gamma\,\mathbb{E}_{Y|D_{s'}}[L[Y,\tilde{f}(x),s]]$, which separates into a clean fair-model loss, a penalty on noisy label transitions, and a penalty on cross-group disagreement in the proxy loss. The paper claims this is why the selected instances are less influenced by label bias, and the experiments support higher accuracy and lower fairness violation than uniform sampling, gradient-norm selection, and RHO-LOSS.

Load-bearing premise

The entire selection score rests on the assumption that the zero-shot predictor $\tilde{f}$ approximates the posterior predictive of a model trained on a clean fair holdout set; if CLIP carries label bias or does not fit the task, the approximated holdout-loss term is biased and the selection can favor the wrong points.

Editorial extensions

If this is right

Training on the selected subset removes the need for a clean holdout set, so the method applies where fair labels are unavailable.
The approach is compatible with any log-likelihood or cross-entropy based classifier and needs no noise-rate estimation, so it can be added to existing training pipelines.
It addresses both label bias and selection bias (via the resampling step), improving demographic parity without sacrificing accuracy in the tested settings.
Because it selects only a fraction of each batch, training converges faster than with uniform sampling or RHO-LOSS, per the reported epochs-to-target-accuracy results.

Reading between the lines

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

If the zero-shot predictor itself carries label bias, the approximated holdout term is biased; the peer-prediction penalty only punishes cross-group disagreement, so a predictor that makes the same confident mistake for both groups would evade the penalty. A sensitivity test swapping in predictors with known group bias would reveal this failure mode.
The decomposition assumes label flips depend on (Z,S) but not on X; under instance-dependent label noise the noisy-loss penalty may misalign, so an experiment with within-group flip rates would test robustness beyond symmetric bias.
The derivation targets demographic parity; equalized odds or calibration would require a different group-conditional penalty, which the same derivation can yield but is not pursued.

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.

Desk Editor's Note

read the letter

The paper puts forward a fairness-aware data selection objective that combines RHO-LOSS, a zero-shot predictor as a holdout proxy, and a peer-prediction fairness penalty. That combination is new, and the resampling step to address selection bias is a sensible addition. The empirical results on CelebA and LFW+a are consistent: the method improves accuracy and reduces demographic parity violation compared to uniform sampling, RHO-LOSS, and gradient-norm baselines, and the ablation across zero-shot predictors and backbones suggests the effect is not tied to one model choice.

The soft spot is not a detail: the central derivation does not hold up. Substituting Eq. (12) into Eq. (11) gives a term of the form L[y|x,D_t,s] − α L[y,f~(x),s] + αγ E[L], with the sum over s still present. The paper instead writes Eq. (13) as L[y|x,D_t,s] + (1−α)L[y,f~(x),s] − γ E[L], dropping the sum and changing both coefficients. Since Eq. (13) is the actual selection function used in training, the method is not what the derivation claims it is. This is a load-bearing error, not a typo in a minor expression.

There are additional concerns. The approximation in Eq. (10) — that the zero-shot predictor's log-posterior equals the posterior expectation of a clean holdout-trained model — is strongly assumed and never validated. The peer-prediction term only penalizes cross-group disagreement; it does not correct the zero-shot predictor's own misclassifications. Experiments are limited to two image datasets, binary labels, and a small hyperparameter grid for α and γ; the paper does not state whether the grid was evaluated on a validation split, so tuning on the test metric is a real possibility. The reproducibility checklist claims code in the appendix, but no code appears in the provided manuscript.

Who should read this? It is most useful to someone working on data selection for fairness who wants a concrete idea for a heuristic: pick training points that have low zero-shot loss plus a cross-group disagreement penalty. That heuristic may be worth exploring, but the theoretical justification as written is wrong.

Recommendation: reject in current form. The authors should redo the derivation (or honestly present Eq. (13) as a heuristic), validate the zero-shot approximation on a held-out clean set, and report a proper hyperparameter-selection protocol. If that happens, the idea deserves another look.

Referee Report

2 major / 5 minor

Summary. The paper proposes a data selection method for fair learning under label bias. It extends the RHO-LOSS criterion of Mindermann et al. (2022) by using a zero-shot predictor (CLIP) as a proxy for a clean holdout model, and augments this proxy with a peer prediction mechanism intended to correct label bias in the selection score. The final selection objective is Eq. (13), which is used in Algorithm 1 to rank training instances. Experiments on CelebA and LFW+a with synthetic symmetric label bias report improved accuracy and lower demographic parity violation relative to uniform sampling, gradient norm selection, and RHO-LOSS, together with ablation studies over the zero-shot backbone, target backbone, and hyperparameters.

Significance. If the derivation were sound, the paper would offer a practical contribution: a modality-agnostic data selection principle that mitigates label bias without requiring a clean holdout set or noise rate estimation. The empirical study is reasonably broad (two datasets, multiple fairness metrics, ablations, convergence analysis) and the reported gains over the baselines are consistent across settings. However, the central theoretical derivation is invalid, and the empirical results cannot by themselves establish the paper's stated claim that Eq. (13) simulates training on a clean holdout set under a fair distribution. The paper also deserves credit for comparing several zero-shot backbones and for reporting selected-data statistics (Fig. 1), which directly probe the method's intended behavior.

major comments (2)

[Fair Data Selection with Peer Prediction Mechanism (Eqs. 11-13)] Eq. (13) does not follow from Eqs. (11) and (12). Substituting Eq. (12) into Eq. (11) for the zero-shot loss term gives Σ_s (C_s/m)[L[y|x,D_t,s] − αL[y,f~(x),s] + αγE_{Y|D_{s'}}[L[Y,f~(x),s]]], not the expression in Eq. (13). The coefficient of the zero-shot term changes from −α to +(1−α), the peer-penalty coefficient changes from αγ to γ, and the sum over s is dropped even though E_{Y|D_{s'}} depends on s through the choice of the opposing group. Since Eq. (13) is the selection criterion used in Algorithm 1 and the paper's stated connection to fair holdout-loss maximization rests on this derivation, the central theoretical claim is unsupported.
[Method, Eq. (10)] Eq. (10) assumes that the zero-shot predictor f~(x) is the posterior mean of a model trained on the clean fair holdout distribution. This is a strong approximation: a pretrained model may carry its own label bias or be misaligned with the task distribution, in which case the 'holdout loss' term in the selection criterion is biased rather than a fair reference. The paper justifies the assumption only by the heuristic argument that the posterior is narrow over a large training set; the ablation over three zero-shot backbones (Table 3) tests sensitivity to the backbone but does not validate the approximation against a true clean holdout model. Because this approximation makes Eq. (9) tractable and underlies the claim that no clean holdout set is needed, it is load-bearing and requires direct validation.

minor comments (5)

[Appendix, derivation of Eq. (12)] In the derivation of the expectation version of the peer loss, the step from the double sum over i and i′≠i to (1/N_s)Σ_i E_{Y|D_{s'}}[γL[Y,f~(x_i),s]] drops the finite-sample factor (N_s−1)/N_s; the equality should be stated as an approximation or include this factor.
[Eq. (14) and appendix Eq. (24)] The notation in Eq. (14) is inconsistent with the appendix: the first term uses L[Y,f~(X),S] while the appendix's Eq. (24) uses L[Z,f~(X),S], and the third term writes L[j,f~(X)] rather than L[j,f~(X),s]; these should be harmonized.
[Experimental setup] The hardware description 'NVIDIA GeForce RTX 3090 with 86GB memory' appears to be a typo, as the RTX 3090 has 24GB of memory.
[Algorithm 1] The loop header 'for t in 0, · · ·, Tdo' contains a typo and should read 'for t in 0, · · ·, T do'.
[Baselines] The implementation details for the baselines, especially RHO-LOSS's use of a holdout set and the hyperparameter settings for the gradient norm variants, are not described in enough detail to allow replication.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the selection objective is anchored to external zero-shot predictions and checked against injected label flips, though the derivation to Eq. (13) contains a non-circular algebraic gap.

full rationale

The central derivation chain is not circular. The selection rule in Eq. (13) is built on external anchors: the RHO-LOSS generalization-loss criterion of Mindermann et al. (2022) and a zero-shot CLIP predictor used as a stand-in for a clean holdout model, following Deng et al. (2023). The paper's fairness claims are checked against known injected label flips (Fig. 1) and held-out accuracy/fairness metrics, so the reported outcomes are not fitted restatements of the method's own inputs. The self-citations, e.g., Zhang et al. (2021) in the paragraph 'Why the fair selection principle pick instances less influenced by label bias?', are not load-bearing: the decomposition of Eq. (12) into Eq. (14) is re-derived in the appendix rather than imported by authority. Two weaknesses are real but are not circularity. First, Eq. (10) assumes the zero-shot predictor's posterior effectively equals the clean-holdout posterior, an unvalidated external assumption; this is a correctness risk, not a circular reduction. Second, the stated transition from Eq. (11) plus Eq. (12) to Eq. (13) is not a valid algebraic substitution: substituting Eq. (12) into Eq. (11) changes the zero-shot coefficient from -alpha to +(1-alpha) and the peer-penalty coefficient from alpha*gamma to gamma, and Eq. (14) decomposes Eq. (12) rather than the final Eq. (13). This is a derivation gap that should be corrected, but it is not a case where a prediction reduces by construction to a fitted parameter or to a self-citation chain.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The method introduces no new physical or conceptual entities. It relies on existing concepts: zero-shot predictors, peer loss, and demographic parity. The main load-bearing assumptions are the fair distribution factorization, the X-conditional independence of label noise, and the adequacy of the zero-shot predictor as a clean holdout proxy. The free parameters α and γ are tuned on validation data.

free parameters (3)

α (scaling factor) = tuned from {0.1, 0.3, 0.5, 0.7, 0.9}
Controls the trade-off between training loss and zero-shot predictor loss in Eq. (13). Selected on validation data without a stated criterion.
γ (peer loss weight) = tuned from {0.1, 0.3, 0.5, 0.7, 0.9}
Controls the penalty for demographic disagreement in the peer-prediction term. Selected on validation data.
Selection ratio Nb/NB = 0.1
Set to match the RHO-LOSS setup, not tuned per dataset.

assumptions (5)

domain assumption Fair distribution factorization p(x,y,s) = p(y|x)p(x)p(s)
Assumed in the Method section to define the ideal fair distribution. It states that in a fair world, labels are independent of sensitive attributes.
domain assumption Label noise flips independent of X given Z and S
Used in the loss decomposition (Eq. 16) where T^s_ij = P(Y=j|Z=i,S=s) is assumed not to depend on X. This simplifies the derivation but may not hold in practice.
ad hoc to paper Zero-shot predictor approximates the posterior mean of the holdout model
Eq. (10) replaces E_{p(θ|D*)} log p(y|x,θ,s) with log p(y|f̃(x),s). Justified by a narrow posterior assumption, but no evidence is provided that CLIP matches the task-specific clean holdout posterior.
standard math Jensen's inequality lower bound is valid and dropping constant terms is safe
The lower bound derivation in the appendix uses Jensen's inequality correctly, and terms independent of (x,y) are dropped, which is standard.
ad hoc to paper Peer prediction mechanism ensures fairness of the zero-shot predictor
The cross-group sampling in Eq. (12) is claimed to eliminate label bias and ensure fairness, but the paper only provides a loss decomposition, not a formal guarantee that the zero-shot predictor becomes fair.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Navigating Towards Fairness with Data Selection." pith.science (2026). https://pith.science/paper/A2PYG2D3

@misc{pith2026241211072,
  author       = {Pith},
  title        = {Pith review of: Navigating Towards Fairness with Data Selection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/A2PYG2D3}},
  note         = {Machine review of arXiv:2412.11072}
}

read the original abstract

Machine learning algorithms often struggle to eliminate inherent data biases, particularly those arising from unreliable labels, which poses a significant challenge in ensuring fairness. Existing fairness techniques that address label bias typically involve modifying models and intervening in the training process, but these lack flexibility for large-scale datasets. To address this limitation, we introduce a data selection method designed to efficiently and flexibly mitigate label bias, tailored to more practical needs. Our approach utilizes a zero-shot predictor as a proxy model that simulates training on a clean holdout set. This strategy, supported by peer predictions, ensures the fairness of the proxy model and eliminates the need for an additional holdout set, which is a common requirement in previous methods. Without altering the classifier's architecture, our modality-agnostic method effectively selects appropriate training data and has proven efficient and effective in handling label bias and improving fairness across diverse datasets in experimental evaluations.

Figures

Figures reproduced from arXiv: 2412.11072 by the authors.

**Figure 2.** Ablation studies on critical hyperparameters, including [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

45 extracted references · 36 canonical work pages

[1]

Barocas, S.; Hardt, M.; and Narayanan, A. 2018. Fairness and Machine Learning Limitations and Opportunities

work page 2018
[2]

Bengio, Y.; Louradour, J.; Collobert, R.; and Weston, J. 2009. Curriculum Learning. In Proceedings of the 26th Annual International Conference on Machine Learning, ICML '09, 41–48. New York, NY, USA: Association for Computing Machinery. ISBN 9781605585161

work page 2009
[3]

Bertrand, M.; and Mullainathan, S. 2004. Are Emily and Greg More Employable Than Lakisha and Jamal? A Field Experiment on Labor Market Discrimination. American Economic Review, 94(4): 991--1013

work page 2004
[4]

Brennan, T.; Dieterich, W.; and Ehret, B. 2009. Evaluating the Predictive Validity of the Compas Risk and Needs Assessment System. Criminal Justice and Behavior, 36(1): 21--40

work page 2009
[5]

Cheng, H.; Zhu, Z.; Li, X.; Gong, Y.; Sun, X.; and Liu, Y. 2021. Learning with Instance-Dependent Label Noise: A Sample Sieve Approach. In International Conference on Learning Representations

work page 2021
[6]

Chowdhury, S. B. R.; and Chaturvedi, S. 2023. Sustaining fairness via incremental learning. In Proceedings of the Thirty-Seventh AAAI Conference on Artificial Intelligence and Thirty-Fifth Conference on Innovative Applications of Artificial Intelligence and Thirteenth Symposium on Educational Advances in Artificial Intelligence, AAAI'23/IAAI'23/EAAI'23. A...

work page 2023
[7]

Coleman, C.; Yeh, C.; Mussmann, S.; Mirzasoleiman, B.; Bailis, P.; Liang, P.; Leskovec, J.; and Zaharia, M. 2019. Selection Via Proxy: Efficient Data Selection For Deep Learning. CoRR, abs/1906.11829

arXiv 2019
[8]

Dai, J. 2020. Label Bias, Label Shift: Fair Machine Learning with Unreliable Labels

work page 2020

Show all 45 references

[9]

Dasgupta, A.; and Ghosh, A. 2013. Crowdsourced judgement elicitation with endogenous proficiency. In Proceedings of the 22nd International Conference on World Wide Web, WWW '13, 319–330. New York, NY, USA: Association for Computing Machinery. ISBN 9781450320351

2013
[10]

Deng, Z.; Cui, P.; and Zhu, J. 2023. Towards Accelerated Model Training via Bayesian Data Selection. In Thirty-seventh Conference on Neural Information Processing Systems

2023
[11]

Dosovitskiy, A.; Beyer, L.; Kolesnikov, A.; Weissenborn, D.; Zhai, X.; Unterthiner, T.; Dehghani, M.; Minderer, M.; Heigold, G.; Gelly, S.; Uszkoreit, J.; and Houlsby, N. 2021. An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale. In International Confere...

2021
[12]

Faliagka, E.; Ramantas, K.; Tsakalidis, A.; and Tzimas, G. 2012. Application of Machine Learning Algorithms to an online Recruitment System

2012
[13]

S.; Shamir, E.; and Tishby, N

Freund, Y.; Seung, H. S.; Shamir, E.; and Tishby, N. 1997. Selective Sampling Using the Query by Committee Algorithm. Mach. Learn., 28(2–3): 133–168

1997
[14]

Hardt, M.; Price, E.; and Srebro, N. 2016. Equality of Opportunity in Supervised Learning. CoRR, abs/1610.02413

2016 arXiv
[15]

He, K.; Zhang, X.; Ren, S.; and Sun, J. 2016. Deep Residual Learning for Image Recognition. In 2016 IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2016, Las Vegas, NV, USA, June 27-30, 2016 , 770--778. IEEE Computer Society

2016
[16]

Huang, G.; Liu, Z.; van der Maaten, L.; and Weinberger, K. Q. 2017. Densely Connected Convolutional Networks. In CVPR, 2261--2269. IEEE Computer Society. ISBN 978-1-5386-0457-1

2017
[17]

H.; Wong, D

Jiang, A. H.; Wong, D. L.; Zhou, G.; Andersen, D. G.; Dean, J.; Ganger, G. R.; Joshi, G.; Kaminsky, M.; Kozuch, M.; Lipton, Z. C.; and Pillai, P. 2019. Accelerating Deep Learning by Focusing on the Biggest Losers. CoRR, abs/1910.00762

2019 arXiv
[18]

Jiang, H.; and Nachum, O. 2020. Identifying and Correcting Label Bias in Machine Learning. In Chiappa, S.; and Calandra, R., eds., Proceedings of the Twenty Third International Conference on Artificial Intelligence and Statistics, volume 108 of Proceedings of Machine Learning ...

2020
[19]

Jiang, Z.; Han, X.; Jin, H.; Wang, G.; Chen, R.; Zou, N.; and Hu, X. 2023. Chasing Fairness Under Distribution Shift: A Model Weight Perturbation Approach. In Thirty-seventh Conference on Neural Information Processing Systems

2023
[20]

Kamiran, F.; and Calders, T. 2012. Data preprocessing techniques for classification without discrimination. Knowledge and Information Systems, 33(1): 1--33

2012
[21]

Katharopoulos, A.; and Fleuret, F. 2018. Not All Samples Are Created Equal: Deep Learning with Importance Sampling. In Dy, J.; and Krause, A., eds., Proceedings of the 35th International Conference on Machine Learning, volume 80 of Proceedings of Machine Learning Research, 252...

2018
[22]

Kawaguchi, K.; and Lu, H. 2019. Ordered SGD: A New Stochastic Optimization Framework for Empirical Risk Minimization. In International Conference on Artificial Intelligence and Statistics

2019
[23]

E.; Kim, A

Khandani, A. E.; Kim, A. J.; and Lo, A. W. 2010. Consumer credit-risk models via machine-learning algorithms. Journal of Banking & Finance, 34(11): 2767 -- 2787

2010
[24]

Killamsetty, K.; Sivasubramanian, D.; Ramakrishnan, G.; of Texas at Dallas, R. I. U.; of Technology Bombay Institution One, I. I.; and Two, I. 2020. GLISTER: Generalization based Data Subset Selection for Efficient and Robust Learning. In AAAI Conference on Artificial Intelligence

2020
[25]

Y.; Yoon, H.; Lee, J.; Kim, N.; and Sung, M.-K

Kim, S.-E.; Paik, H. Y.; Yoon, H.; Lee, J.; Kim, N.; and Sung, M.-K. 2015. Sex- and gender-specific disparities in colorectal cancer risk. World journal of gastroenterology : WJG, 21: 5167--5175

2015
[26]

Konstantinov, N.; and Lampert, C. H. 2022. Fairness-Aware PAC Learning from Corrupted Data. Journal of Machine Learning Research, 23(160): 1--60

2022
[27]

J.; Jung, J.; Goel, S.; and Skeem, J

Lin, Z. J.; Jung, J.; Goel, S.; and Skeem, J. 2020. The limits of human predictions of recidivism. Science Advances, 6(7): eaaz0652

2020
[28]

Liu, Y.; and Guo, H. 2020. Peer Loss Functions: Learning from Noisy Labels without Knowing Noise Rates. In Proceedings of the 37th International Conference on Machine Learning, ICML 2020, 13-18 July 2020, Virtual Event , volume 119 of Proceedings of Machine Learning Research, ...

2020
[29]

Liu, Z.; Luo, P.; Wang, X.; and Tang, X. 2015. Deep Learning Face Attributes in the Wild. In 2015 IEEE International Conference on Computer Vision (ICCV), 3730--3738

2015
[30]

Loshchilov, I.; and Hutter, F. 2015. Online Batch Selection for Faster Training of Neural Networks. CoRR, abs/1511.06343

2015 arXiv
[31]

Loshchilov, I.; and Hutter, F. 2017. Decoupled Weight Decay Regularization. In International Conference on Learning Representations

2017
[32]

M.; Razzak, M

Mindermann, S.; Brauner, J. M.; Razzak, M. T.; Sharma, M.; Kirsch, A.; Xu, W.; H \"o ltgen, B.; Gomez, A. N.; Morisot, A.; Farquhar, S.; and Gal, Y. 2022. Prioritized Training on Points that are Learnable, Worth Learning, and not yet Learnt. In Chaudhuri, K.; Jegelka, S.; Song...

2022
[33]

Moore, J. S. 1998. An Expert System Approach to Graduate School Admission Decisions and Academic Performance Prediction . Omega, 26(5): 659--670

1998
[34]

Mukerjee, A.; Biswas, R.; Kalyanmoy, Y.; Amrit, D.; and Mathur, P. 2002. Multi-objective Evolutionary Algorithms for the Risk-return Trade-off in Bank Loan Management. International Transactions in Operational Research, 9

2002
[35]

O.; Alabdulmohsin, I.; Schnider, E.; Opsahl-Ong, K.; Brown, A.; Roy, S.; Mincu, D.; Chen, C.; Dieng, A.; Liu, Y.; Natarajan, V.; Karthikesalingam, A.; Heller, K

Schrouff, J.; Harris, N.; Koyejo, O. O.; Alabdulmohsin, I.; Schnider, E.; Opsahl-Ong, K.; Brown, A.; Roy, S.; Mincu, D.; Chen, C.; Dieng, A.; Liu, Y.; Natarajan, V.; Karthikesalingam, A.; Heller, K. A.; Chiappa, S.; and D'Amour, A. 2022. Diagnosing failures of fairness transfe...

2022
[36]

S.; Opper, M.; and Sompolinsky, H

Seung, H. S.; Opper, M.; and Sompolinsky, H. 1992. Query by committee. In Proceedings of the Fifth Annual Workshop on Computational Learning Theory, COLT '92, 287–294. New York, NY, USA: Association for Computing Machinery. ISBN 089791497X

1992
[37]

Shnayder, V.; Agarwal, A.; Frongillo, R.; and Parkes, D. C. 2016. Informed Truthfulness in Multi-Task Peer Prediction. In Proceedings of the 2016 ACM Conference on Economics and Computation, EC '16, 179–196. New York, NY, USA: Association for Computing Machinery. ISBN 9781450339360

2016
[38]

Wang, J.; Liu, Y.; and Levy, C. 2021. Fair Classification with Group-Dependent Label Noise. In Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency, FAccT '21, 526–536. New York, NY, USA: Association for Computing Machinery. ISBN 9781450383097

2021
[39]

Wick, M.; panda, s.; and Tristan, J.-B. 2019. Unlocking Fairness: A Trade-off Revisited. In Wallach, H.; Larochelle, H.; Beygelzimer, A.; d Alch\' e -Buc, F.; Fox, E.; and Garnett, R., eds., Advances in Neural Information Processing Systems, volume 32, 8783--8792. Curran Assoc...

2019
[40]

Wolf, L.; Hassner, T.; and Taigman, Y. 2011. Effective Unconstrained Face Recognition by Combining Multiple Descriptors and Learned Background Statistics . IEEE Transactions on Pattern Analysis and Machine Intelligence, 33(10): 1978--1990

2011
[41]

Yeh, I.; and Lien, C.-H. 2009. The comparisons of data mining techniques for the predictive accuracy of probability of default of credit card clients. Expert Systems with Applications, 36: 2473--2480

2009
[42]

Zhang, Y.; Zhou, F.; Li, Z.; Wang, Y.; and Chen, F. 2021. Bias-tolerant Fair Classification. In Balasubramanian, V. N.; and Tsang, I., eds., Proceedings of The 13th Asian Conference on Machine Learning, volume 157 of Proceedings of Machine Learning Research, 840--855. PMLR

2021
[43]

Zhang, Y.; Zhou, F.; Li, Z.; Wang, Y.; and Chen, F. 2023. Fair Representation Learning with Unreliable Labels. In Ruiz, F.; Dy, J.; and van de Meent, J.-W., eds., Proceedings of The 26th International Conference on Artificial Intelligence and Statistics, volume 206 of Proceedi...

2023
[44]

, " * write output.state after.block = add.period write newline

ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all...
[45]

write newline

" write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.

[1] [1]

Barocas, S.; Hardt, M.; and Narayanan, A. 2018. Fairness and Machine Learning Limitations and Opportunities

work page 2018

[2] [2]

Bengio, Y.; Louradour, J.; Collobert, R.; and Weston, J. 2009. Curriculum Learning. In Proceedings of the 26th Annual International Conference on Machine Learning, ICML '09, 41–48. New York, NY, USA: Association for Computing Machinery. ISBN 9781605585161

work page 2009

[3] [3]

Bertrand, M.; and Mullainathan, S. 2004. Are Emily and Greg More Employable Than Lakisha and Jamal? A Field Experiment on Labor Market Discrimination. American Economic Review, 94(4): 991--1013

work page 2004

[4] [4]

Brennan, T.; Dieterich, W.; and Ehret, B. 2009. Evaluating the Predictive Validity of the Compas Risk and Needs Assessment System. Criminal Justice and Behavior, 36(1): 21--40

work page 2009

[5] [5]

Cheng, H.; Zhu, Z.; Li, X.; Gong, Y.; Sun, X.; and Liu, Y. 2021. Learning with Instance-Dependent Label Noise: A Sample Sieve Approach. In International Conference on Learning Representations

work page 2021

[6] [6]

Chowdhury, S. B. R.; and Chaturvedi, S. 2023. Sustaining fairness via incremental learning. In Proceedings of the Thirty-Seventh AAAI Conference on Artificial Intelligence and Thirty-Fifth Conference on Innovative Applications of Artificial Intelligence and Thirteenth Symposium on Educational Advances in Artificial Intelligence, AAAI'23/IAAI'23/EAAI'23. A...

work page 2023

[7] [7]

Coleman, C.; Yeh, C.; Mussmann, S.; Mirzasoleiman, B.; Bailis, P.; Liang, P.; Leskovec, J.; and Zaharia, M. 2019. Selection Via Proxy: Efficient Data Selection For Deep Learning. CoRR, abs/1906.11829

arXiv 2019

[8] [8]

Dai, J. 2020. Label Bias, Label Shift: Fair Machine Learning with Unreliable Labels

work page 2020