Pith. sign in

REVIEW 4 major objections 6 minor 46 references

DeepIFSAC: Deep Imputation of Missing Values Using Feature and Sample Attention within Contrastive Framework

T0 review · 4 major / 6 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 A joint feature-sample attention imputer claims top rank among 11 missing-value methods.

desk verdict DeepIFSAC is a solid incremental study whose central superiority claim is not backed by significance tests; the claim should be softened. read the letter →

arxiv 2501.10910 v3 pith:NDAVNQEC submitted 2025-01-19 cs.LG stat.ML

Ibna Kowsar , Shourav B. Rabbani , Yina Hou , Manar D. Samad This is my paper · ORCID
classification cs.LGstat.ML
keywords missingvalueimputationtabulardataattentionmechanismcontrastivelearningCutMixaugmentationMNARelectronichealthrecordsNRMSE
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

Missing values in real-world tabular data are often neither few nor random, and standard imputation methods degrade as the missing rate climbs. DeepIFSAC is a deep imputation model that estimates a missing cell by attending both to other features in the same row and to similar samples in the same dataset, and it trains these two attention passes together with CutMix corruption and a contrastive loss. The paper's central claim is that this combined framework achieves the best average performance rank among eleven statistical, machine-learning, and deep baselines across twelve tabular data sets, for missing rates from 10% to 90% and for the three standard missingness mechanisms—MCAR (missing completely at random), MAR (missing at random), and MNAR (missing not at random)—with the clearest advantage under MNAR. A downstream test on real electronic health records reports the highest F1 for heart-failure classification when the features are imputed by DeepIFSAC. If the claim holds, the practical message is that between-sample attention and augmentation with real values can make deep imputation robust in exactly the high-rate, non-random regimes where classical methods struggle.

What carries the argument

The central object is a two-stage multi-head self-attention pipeline over a per-feature embedding tensor. Stage one applies multi-head self-attention across features to capture column dependencies; stage two reshapes the embedding and applies multi-head self-attention across samples to capture row dependencies, after which an MLP head reconstructs the data values. Training corrupts mini-batches with CutMix, replacing a random subset of a sample's features with the corresponding values from another sample, and optimizes the sum of a reconstruction loss computed only on observed entries and an infoNCE contrastive loss that pulls corrupted and uncorrupted embeddings of the same sample together. The CutMix corruption, rather than zero-filling, is what lets the model learn to repair missing-like patterns from real value combinations, while contrastive learning sharpens the sample neighbourhoods that the between-sample attention then relies on.

What would settle it

Take a dataset with an MNAR mechanism whose missingness propensity can be estimated, and train two versions of DeepIFSAC: one with standard CutMix corruption and one with corruption masks drawn from the true propensity. If the true-propensity model does not achieve lower NRMSE on held-out missing entries, then the CutMix surrogate for non-random missingness is not doing the work the paper assigns it.

Watch

Extended reading notes

Core claim

The paper's central claim is that a missing value, lying at the intersection of a row and a column, should be reconstructed by jointly learning between-feature attention and between-sample attention rather than by either alone. In the reported experiments, DeepIFSAC is compared with eleven baselines on twelve tabular data sets under three simulated missingness mechanisms and five missing rates, with methods ranked by normalized root mean squared error on held-out folds. The average rank places DeepIFSAC first overall for MCAR, MAR, and MNAR, and its advantage over the strongest alternatives is largest under MNAR, where missingness depends on the value itself and between-sample attention can supply information that between-feature attention cannot. On real electronic health records without ground truth, the imputed data support a heart-failure classifier with the highest mean F1 score among the compared methods. The paper additionally reports ablations showing that CutMix augmentation improves on zero-filling and that the contrastive loss improves on joint attention without it.

Load-bearing premise

The load-bearing premise is that CutMix corruption—replacing random features with values from another sample—is a faithful stand-in for real missingness, including MNAR where the chance of being missing depends on the missing value itself; if the two distributions diverge, the model could learn to repair CutMix patterns while producing biased imputations for true missing cells.

Editorial extensions

If this is right

  • If the average-rank result holds, DeepIFSAC is the best default among the eleven tested methods for tabular data with missing rates from 10% to 90% across MCAR, MAR, and MNAR.
  • The MNAR result implies that row attention can partially compensate for the absence of feature dependencies, the regime where regression-based and generative baselines weaken.
  • The EHR result implies that lower imputation error translates into better predictive performance on real data where missing ground-truth values are unknowable.
  • The ablations imply that CutMix augmentation with real values and the contrastive objective are load-bearing parts of the method, not optional extras.
  • At high missing rates, between-sample attention is reported as the most resilient view, which supports attention-based imputation as the stronger deep-learning family for tabular missing values.

Reading between the lines

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

  • The paper leaves the CutMix-to-MNAR transfer unvalidated; a direct test would train on masks drawn from an estimated missingness propensity and compare NRMSE against CutMix-trained models, showing whether the surrogate assumption holds.
  • The dataset heterogeneity suggests the practical rule is conditional, not universal: attention-based imputation appears strongest with adequate sample sizes and comparable feature scales, while tree-based methods remain competitive on small or extreme-scale datasets; a formal decision rule could make this actionable.
  • Contrastive learning probably helps most when samples form meaningful clusters, so on weakly clustered datasets the between-sample attention could dilute rather than help; this is testable by varying clusterability.
  • Extending the reconstruction head to categorical features with a classifier loss is a natural next step, and the same attention weights could double as a diagnostic for which features and samples drive the imputation.
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

4 major / 6 minor

Summary. The paper proposes DeepIFSAC, a transformer-based missing value imputation method that combines between-feature (column) attention and between-sample (row) attention, trained with CutMix data augmentation and a contrastive learning objective. The method is benchmarked against 11 baselines on 12 OpenML tabular datasets under MCAR, MAR, and MNAR missingness at rates from 10% to 90%, using held-out test folds and NRMSE as the primary metric; a secondary evaluation imputes missing values in an All of Us EHR dataset and measures downstream heart-failure classification F1. The central claim is that DeepIFSAC's average performance rank across datasets demonstrates superiority over state-of-the-art methods, especially under MNAR.

Significance. If the central claim were established, the paper would provide a valuable practical result: a relatively simple attention-based imputer that outperforms GAN-, diffusion-, and autoencoder-based methods across a wide range of missing-data scenarios, with a large and careful benchmark design. The evaluation has notable strengths: set-aside test folds avoid circular fitting, the missingness regimes cover three mechanisms, the missing rates extend to 90%, and a real EHR downstream task is included. However, the statistical support for the superiority claim is currently insufficient: the margins over the closest baseline are small relative to the reported variability, no significance tests are provided, and the acknowledged shorter training budget for generative baselines hampers fair comparison. The novelty of the contrastive component is also undermined by the ablation results, which show essentially no improvement over the same architecture without contrastive learning.

major comments (4)
  1. [Section 5.1, Table 2] The abstract's claim of superiority for missing rates between 10% and 90% is not supported by the reported average ranks. The gap between DeepIFSAC and the closest baseline (Between-sample [40]) is 0.19 (MCAR), 0.33 (MAR), and 0.23 (MNAR) rank points, while the standard deviations of the ranks are roughly 2.5–3.5; no significance test, confidence interval, or effect size is reported. Moreover, at MCAR 70% (4.33 vs. 3.75) and MAR 70% (4.67 vs. 4.33) the Between-sample baseline has a better average rank than DeepIFSAC, directly contradicting the phrase 'for missing rates between 10% and 90%.' The authors should either provide paired tests across datasets (e.g., Wilcoxon signed-rank) or substantially temper the superiority claim.
  2. [Section 3.4, Table 8] The contrastive learning component is a central novel contribution, but the ablation shows that it provides at most 0.01 NRMSE improvement over the same model without the contrastive loss (e.g., dataset 37 at 30%: 0.30 vs. 0.29; dataset 1497 at 10%: 0.18 vs. 0.18). Given the standard deviations reported in the same table (on the order of 0.01–0.06), this difference is within noise. The authors should demonstrate that the contrastive contribution is statistically significant, or reposition contrastive learning as a minor auxiliary loss rather than a central novelty.
  3. [Section 4.4 and Section 6.1] The deep baselines GAIN and Diffputer are trained for 1,000 epochs, while the authors themselves acknowledge in Section 6.1 that these methods are typically trained for 10,000 epochs. This comparison likely understates the performance of the generative baselines and undermines the claim that DeepIFSAC is superior to state-of-the-art deep methods. To support the central claim, the authors should retrain these baselines to convergence (or at least match their recommended training budgets) or provide evidence that 1,000 epochs is sufficient for them.
  4. [Section 3.2, Algorithm 1] The training procedure uses CutMix corruption (Equation 4) as the sole surrogate for missingness, including for MNAR where the missingness probability depends on the missing value itself. The paper does not justify that CutMix corruptions are a faithful surrogate for the target missingness mechanisms; the empirical NRMSE is the evidence offered, but that is the very claim under evaluation. The authors should either provide a diagnostic separating the effect of the augmentation distribution from the true missingness model or explicitly discuss this limitation as a threat to generalization.
minor comments (6)
  1. [Table 6] The downstream F1 difference between DeepIFSAC and Between-sample is only 0.25 points (87.49 vs. 87.24) with overlapping standard deviations (1.35 vs. 1.24); no significance test is reported for this claim either.
  2. [Footnote 1] The text states the source code has been publicly shared, but the footnote says the link would be shared in the final version for anonymity; this inconsistency should be resolved and the code link provided.
  3. [References] References [43] and [44] appear to duplicate the same paper; please check and remove the duplicate.
  4. [Equation 14] The NRMSE formula uses Var(X_j), but when X_j has missing values it is unclear whether the variance is computed on observed entries only or after some initialization; please clarify.
  5. [Section 4.3] The two attention baselines 'Between-feature' and 'Between-sample' are from the authors' own prior work [40]; this should be explicitly acknowledged as a limitation of the independence of the comparison.
  6. [Table 2] The average-rank rows would be easier to interpret if accompanied by confidence intervals or error bars, since the raw standard deviations are large and the visual figures in the appendix do not display them.

Circularity Check

0 steps flagged · score 2.0 of 10

No circular derivation: DeepIFSAC is benchmarked on held-out folds against external baselines, and the same-author citations are not load-bearing.

full rationale

The paper's central imputation claim is tested by training on four folds and computing NRMSE on a left-out test fold with simulated missing values (Section 4.5, Equation 14), so no target missing values are fitted and the reconstruction loss (Equations 11-12) is evaluated only on observed entries. The main comparison is against external methods (MICE, missForest, GAIN, Diffputer, DSAN, MIDASpy, KNN, median), whose results are computed in this paper rather than imported from a citation, giving independent grounding. The only same-author citations are [40], used to name the between-feature and between-sample attention ablations, and [29], cited alongside external works [30, 32] for CutMix benefits; neither citation supplies the numerical superiority claim, because those baselines are re-implemented and re-scored here. The CutMix surrogate for MNAR missingness is an assumption that may limit validity, but it is not circular: the evaluation missingness is generated separately from the corruption used in training. The paper's own limitation notes, such as training GAIN and Diffputer for 1,000 instead of their original 10,000 epochs (Section 6.1), weaken the comparative claim but do not make it circular. The absence of significance testing on average-rank margins is a statistical-support concern, not a circularity concern. Overall the derivation is self-contained, with only a minor same-author citation burden that does not carry the argument.

Assumptions & free parameters 7 free parameters · 4 assumptions · 0 invented entities

The method uses existing building blocks and introduces no new theoretical entities. The main unstated ingredients are the CutMix surrogate assumption and several unreported hyperparameters, which are the largest burden on the reader.

free parameters (7)
  • CutMix corruption rate
    The fraction of feature values replaced in Equation 4 is never stated; it is a hidden knob that affects training and results.
  • Contrastive temperature tau
    Equation 13 contains tau but no value is reported, so the strength of the contrastive pull is unspecified.
  • Embedding size d = 32
    Taken from SAINT (ref [30]), not fitted to the target results.
  • Number of attention heads H = 8
    Taken from SAINT (ref [30]).
  • Number of transformer blocks L = 6
    Taken from SAINT (ref [30]).
  • Learning rate = 0.0001
    Fixed across models, Section 4.4.
  • Training epochs = 1000
    Fixed across models; shorter than the 10,000 epochs used in GAIN and Diffputer original papers, which may handicap those baselines.
assumptions (4)
  • domain assumption CutMix corruption is a valid surrogate for MCAR, MAR, and MNAR missingness mechanisms
    Section 3.2 and 3.3.3 use CutMix corrupted views to train reconstruction; no theoretical or diagnostic justification that this transfer holds.
  • domain assumption The synthetic missingness generators produce realistic MCAR, MAR, and MNAR patterns
    Section 4 states three types are simulated but never specifies the exact generation mechanism or parameters, so the external validity of the benchmark assumes these generators are standard.
  • domain assumption Between-sample attention over the full test batch is valid at inference
    Section 3.3.2 applies MHSA over the b samples; for test folds this means attention over the whole left-out fold, which may not be available in online imputation settings.
  • standard math Standard transformer softmax attention and infoNCE loss are well-defined
    Equations 8 and 13 rely on standard machinery from Vaswani et al. and Chen et al.; treated as accepted background.

how reviews work

0 comments
Cite this review

Pith. "Pith review of DeepIFSAC: Deep Imputation of Missing Values Using Feature and Sample Attention within Contrastive Framework." pith.science (2026). https://pith.science/paper/NDAVNQEC

@misc{pith2026250110910,
  author       = {Pith},
  title        = {Pith review of: DeepIFSAC: Deep Imputation of Missing Values Using Feature and Sample Attention within Contrastive Framework},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NDAVNQEC}},
  note         = {Machine review of arXiv:2501.10910}
}
read the original abstract

Missing values of varying patterns and rates in real-world tabular data pose a significant challenge in developing reliable data-driven models. The most commonly used statistical and machine learning methods for missing value imputation may be ineffective when the missing rate is high and not random. This paper explores row and column attention in tabular data as between-feature and between-sample attention in a novel framework to reconstruct missing values. The proposed method uses CutMix data augmentation within a contrastive learning framework to improve the uncertainty of missing value estimation. The performance and generalizability of trained imputation models are evaluated in set-aside test data folds with missing values. The proposed framework is compared with 11 state-of-the-art statistical, machine learning, and deep imputation methods using 12 diverse tabular data sets. The average performance rank of our proposed method demonstrates its superiority over the state-of-the-art methods for missing rates between 10% and 90% and three missing value types, especially when the missing values are not random. The quality of the imputed data using our proposed method is compared in a downstream patient classification task using real-world electronic health records. This paper highlights the heterogeneity of tabular data sets to recommend imputation methods based on missing value types and data characteristics.

Figures

Figures reproduced from arXiv: 2501.10910 by the authors.

Figure 1
Figure 1. Proposed missing value imputation framework. Between-feature attention learning is followed by between [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Data splits for evaluating imputation model performance and quality of imputed data in classification tasks. [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Average rank order of individual imputation methods obtained across 10% to 90% MCAR type values. [PITH_FULL_IMAGE:figures/full_fig_p017_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Average rank order of individual imputation methods obtained across 10% to 90% MAR type values. [PITH_FULL_IMAGE:figures/full_fig_p017_4.png]
Figure 5
Figure 5. Figure 5: Average rank order of individual imputation methods obtained across 10% to 90% MNAR type values. [PITH_FULL_IMAGE:figures/full_fig_p018_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

46 extracted references · 33 canonical work pages

  1. [40]

    Kowsar, S

    I. Kowsar, S. B. Rabbani, M. D. Samad, Attention-based imputation of missing values in electronic health records tabular data, in: The 12th IEEE International Conference on Healthcare Informatics (ICHI), 2024

    work page 2024

  2. [1]

    Kazijevs, M

    M. Kazijevs, M. D. Samad, Deep imputation of missing values in time series health data: A review with bench- marking, Journal of biomedical informatics 144 (2023) 104440. doi:10.1016/j.jbi.2023.104440

    work page arXiv 2023

  3. [2]

    Grinsztajn, E

    L. Grinsztajn, E. Oyallon, G. Varoquaux, Why do tree-based models still outperform deep learning on typical tabular data?, in: Thirty-sixth Conference on Neural Information Processing Systems Datasets and Benchmarks Track, 2022

    2022

  4. [3]

    Borisov, T

    V . Borisov, T. Leemann, K. Seßler, J. Haug, M. Pawelczyk, G. Kasneci, Deep neural networks and tabular data: A survey, IEEE Transactions on Neural Networks and Learning Systems (2022)

    work page 2022

  5. [4]

    Kadra, M

    A. Kadra, M. Lindauer, F. Hutter, J. Grabocka, Well-tuned simple nets excel on tabular datasets, Advances in neural information processing systems 34 (2021) 23928–23941

    work page 2021

  6. [5]

    Gorishniy, I

    Y . Gorishniy, I. Rubachev, V . Khrulkov, A. Babenko, Revisiting deep learning models for tabular data, Advances in Neural Information Processing Systems 23 (2021) 18932–18943

    work page 2021

  7. [6]

    Resche-Rigon, I

    M. Resche-Rigon, I. R. White, Multiple imputation by chained equations for systematically and sporadically missing multilevel data, Statistical methods in medical research 27 (6) (2018) 1634

    work page 2018

  8. [7]

    D. J. Stekhoven, P. B¨uhlmann, MissForest—non-parametric missing value imputation for mixed-type data, Bioin- formatics 28 (2012) 112–118

    work page 2012

Show all 46 references
  1. [8]

    H. Khan, M. T. Rasheed, H. Liu, S. Zhang, High-order polynomial interpolation with cnn: A robust approach for missing data imputation, Computers and Electrical Engineering 119 (2024) 109524

    2024

  2. [9]

    H. Khan, X. Wang, H. Liu, Handling missing data through deep convolutional neural network, Information Sciences 595 (2022) 278–293

    2022

  3. [10]

    R. Wu, A. Zhang, I. Ilyas, T. Rekatsinas, Attention-based learning for missing data imputation in holoclean, Proceedings of Machine Learning and Systems 2 (2020) 307–325

    2020

  4. [11]

    Lee, H.-j

    D.-H. Lee, H.-j. Kim, A self-attention-based imputation technique for enhancing tabular data quality, Data 8 (6) (2023) 102

    2023

  5. [12]

    J. Yoon, J. Jordon, M. Schaar, Gain: Missing data imputation using generative adversarial nets, in: International conference on machine learning, PMLR, 2018, pp. 5689–5698

    2018

  6. [13]

    Ouyang, L

    Y . Ouyang, L. Xie, C. Li, G. Cheng, MissDiff: Training diffusion models on tabular data with missing values, in: ICML 2023 Workshop on Structured Probabilistic Inference & Generative Modeling, 2023

    2023

  7. [14]

    S. Yun, D. Han, S. J. Oh, S. Chun, J. Choe, Y . Yoo, CutMix: Regularization strategy to train strong classifiers with localizable features, in: Proceedings of the IEEE/CVF international conference on computer vision, 2019, pp. 6023–6032

    2019

  8. [15]

    R. Lall, T. Robinson, The MIDAS Touch: Accurate and scalable missing-data imputation with deep learning, Political Analysis 30 (2) (2022) 179–196. doi:10.1017/pan.2020.49

    2022 doi

  9. [16]

    M. D. Samad, S. Abrar, N. Diawara, Missing value estimation using clustering and deep learning within multiple imputation framework, Knowledge-based systems 249 (2022) 108968

    2022

  10. [17]

    Bertsimas, C

    D. Bertsimas, C. Pawlowski, Y . D. Zhuo, From predictive methods to missing data imputation: an optimization approach, Journal of Machine Learning Research 18 (196) (2018) 1–39. 14 A PREPRINT - MARCH 26, 2025

    2018

  11. [18]

    Emmanuel, T

    T. Emmanuel, T. Maupong, D. Mpoeleng, T. Semong, B. Mphago, O. Tabona, A survey on missing data in machine learning, Journal of Big data 8 (2021) 1–37

    2021

  12. [19]

    Mera-Gaona, U

    M. Mera-Gaona, U. Neumann, R. Vargas-Canas, D. M. L ´opez, Evaluating the impact of multivariate imputation by mice in feature selection, Plos one 16 (7) (2021) e0254720

    2021

  13. [20]

    J. Chen, J. Yan, D. Z. Chen, J. Wu, Excelformer: A neural network surpassing gbdts on tabular data, arXiv preprint arXiv:2301.02819 (1 2023). URL http://arxiv.org/abs/2301.02819

    2023 arXiv

  14. [21]

    Z. Chen, H. Li, F. Wang, O. Zhang, H. Xu, X. Jiang, Z. Song, H. Wang, Rethinking the diffusion models for missing data imputation: A gradient flow perspective, in: The Thirty-eighth Annual Conference on Neural Information Processing Systems, 2024. URL https://openreview.net/fo...

    2024

  15. [22]

    Y . Wen, Y . Wang, K. Yi, J. Ke, Y . Shen, DiffImpute: Tabular data imputation with denoising diffusion probabilis- tic model, in: 2024 IEEE International Conference on Multimedia and Expo (ICME), IEEE Computer Society, Los Alamitos, CA, USA, 2024, pp. 1–6. doi:10.1109/ICME575...

    2024

  16. [23]

    Y . Sun, J. Li, Y . Xu, T. Zhang, X. Wang, Deep learning versus conventional methods for missing data imputation: A review and comparative study, Expert Systems with Applications 227 (2023) 120201

    2023

  17. [24]

    Jarrett, B

    D. Jarrett, B. C. Cebere, T. Liu, A. Curth, M. van der Schaar, HyperImpute: Generalized iterative imputation with automatic model selection, in: International Conference on Machine Learning, PMLR, 2022, pp. 9916–9937

    2022

  18. [25]

    Liguori, R

    A. Liguori, R. Markovic, M. Ferrando, J. Frisch, F. Causone, C. van Treeck, Augmenting energy time-series for data-efficient imputation of missing values, Applied Energy 334 (2023) 120701

    2023

  19. [26]

    Gondara, K

    L. Gondara, K. Wang, MIDA: Multiple imputation using denoising autoencoders, in: Advances in Knowl- edge Discovery and Data Mining: 22nd Pacific-Asia Conference, PAKDD 2018, Melbourne, VIC, Aus- tralia, June 3-6, 2018, Proceedings, Part III, Springer-Verlag, Berlin, Heidelberg...

    2018 doi

  20. [27]

    L. Zhao, Z. Chen, Z. Yang, Y . Hu, M. S. Obaidat, Local similarity imputation based on fast clustering for incomplete data in cyber-physical systems, IEEE Systems Journal 12 (2018) 1610–1620

    2018

  21. [28]

    Kossen, N

    J. Kossen, N. Band, C. Lyle, A. N. Gomez, T. Rainforth, Y . Gal, Self-attention between datapoints: Going beyond individual input-output pairs in deep learning, Advances in Neural Information Processing Systems 34 (2021) 28742–28756

    2021

  22. [29]

    S. B. Rabbani, I. V . Medri, M. D. Samad, Attention versus contrastive learning of tabular data: a data-centric benchmarking, International Journal of Data Science and Analytics (2024) 1–23

    2024

  23. [30]

    Somepalli, A

    G. Somepalli, A. Schwarzschild, M. Goldblum, C. B. Bruss, T. Goldstein, Saint: Improved neural networks for tabular data via row attention and contrastive pre-training, in: NeurIPS 2022 First Table Representation Workshop, 2022

    2022

  24. [31]

    H. Khan, X. Wang, H. Liu, Missing value imputation through shorter interval selection driven by fuzzy c-means clustering, Computers & Electrical Engineering 93 (2021) 107230

    2021

  25. [32]

    Bahri, H

    D. Bahri, H. Jiang, Y . Tay, D. Metzler, SCARF: Self-supervised contrastive learning using random feature cor- ruption, in: International Conference on Learning Representations, 2022. URL https://openreview.net/forum?id=CuV_qYkmKb3

    2022

  26. [33]

    Vaswani, N

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, Ł. Kaiser, I. Polosukhin, Attention is all you need, Advances in neural information processing systems 30 (2017)

    2017

  27. [34]

    K. He, X. Zhang, S. Ren, J. Sun, Deep residual learning for image recognition, in: Proceedings of the IEEE conference on computer vision and pattern recognition, 2016, pp. 770–778

    2016

  28. [35]

    T. Chen, S. Kornblith, M. Norouzi, G. Hinton, A simple framework for contrastive learning of visual represen- tations, in: H. D. III, A. Singh (Eds.), Proceedings of the 37th International Conference on Machine Learning, V ol. 119 of Proceedings of Machine Learning Research, P...

    2020

  29. [36]

    of Us Research Program Investigators, J

    A. of Us Research Program Investigators, J. C. Denny, J. L. Rutter, D. B. Goldstein, A. Philippakis, J. W. Smoller, G. Jenkins, E. Dishman, The ”all of us” research program, New England Journal of Medicine 381 (2019) 668–

    2019

  30. [37]

    Vanschoren, J

    J. Vanschoren, J. N. van Rijn, B. Bischl, L. Torgo, OpenML: networked science in machine learning, SIGKDD Explorations 15 (2) (2013) 49–60. doi:10.1145/2641190.2641198

    2013

  31. [38]

    S. G. Liao, Y . Lin, D. D. Kang, D. Chandra, J. Bon, N. Kaminski, F. C. Sciurba, G. C. Tseng, Missing value imputation in high-dimensional phenomic data: imputable or not, and how?, BMC bioinformatics 15 (1) (2014) 1–12

    2014

  32. [39]

    Zhang, L

    H. Zhang, L. Fang, P. S. Yu, Unleashing the potential of diffusion models for incomplete data imputation, in: International Conference on Learning Representations, 2024

    2024

  33. [41]

    M. D. Samad, A. Ulloa, G. J. Wehner, L. Jing, D. Hartzel, C. W. Good, B. A. Williams, C. M. Haggerty, B. K. Fornwalt, Predicting survival from large echocardiography and electronic health record datasets: Optimization with machine learning, JACC: Cardiovascular Imaging 12 (201...

    2019 doi

  34. [42]

    Zheng, N

    S. Zheng, N. Charoenphakdee, Diffusion models for missing value imputation in tabular data, in: NeurIPS 2022 First Table Representation Workshop, 2022. URL https://openreview.net/forum?id=4q9kFrXC2Ae

    2022

  35. [43]

    L. Li, Y . Fang, J. Wu, J. Wang, Y . Ge, Encoder–decoder full residual deep networks for robust regression and spatiotemporal estimation, IEEE Transactions on Neural Networks and Learning Systems 32 (9) (2021) 4217–

    2021

  36. [44]

    L. Li, Y . Fang, J. Wu, J. Wang, Y . Ge, Encoder-decoder full residual deep networks for robust regression and spatiotemporal estimation, IEEE Transactions on Neural Networks and Learning Systems 32 (9) (2021) 4217–

    2021

  37. [47]

    doi:10.1109/TNNLS.2020.3017200. 16 A PREPRINT - MARCH 26, 2025 A Additional Experimental Results This section provides additional results demonstrating the average rank order of all imputation methods for the three missing data types. Missing Completely at Random (MCAR) in Fig...

    2020

  38. [676]

    doi:10.1056/NEJMsr1809937

    doi

Pith tools

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