Pith. sign in

REVIEW 3 major objections 5 minor 32 references

Imputation of Missing Data in Smooth Pursuit Eye Movements Using a Self-Attention-based Deep Learning Approach

T0 review · 3 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read By combining a self-attention imputer with a refinement autoencoder, this paper reconstructs missing segments of smooth pursuit eye movement recordings with lower time-domain error than PCHIP, SSA, and KNN, while keeping spectral content…

desk verdict A useful, incremental application paper on SPEM imputation that is worth reviewing but currently overclaims and needs error bars and a fix to its missingness generation. read the letter →

arxiv 2506.00545 v1 pith:2SU2DA2R submitted 2025-05-31 cs.LG cs.AI

Mehdi Bejani , Guillermo Perez-de-Arenaza-Pozo , Julián D. Arias-Londoño , Juan I. Godino-LLorente This is my paper
classification cs.LGcs.AI
keywords missingdataimputationsmoothpursuiteyemovementsself-attentionSAITSrefinementautoencoderblinkartifactsParkinson'sdiseasebiomedicaltimeseries
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 deep-learning imputation pipeline can reconstruct missing segments in smooth pursuit eye movement recordings accurately enough to make the repaired signals usable for clinical analysis. The pipeline first imputes downsampled sequences with SAITS, a self-attention-based imputation network, then restores full resolution by cubic interpolation, and finally passes the signal through a custom convolutional autoencoder trained only on complete sequences. On 5,504 recordings from Parkinsonian patients and healthy controls, the authors report that this SAITS-RAE pipeline lowers mean absolute error, mean relative error, and root mean square error relative to PCHIP, SSA, and KNN while keeping frequency-domain error low. The advantage grows when entire four-second intervals are missing, which is the case where classical interpolation degrades most. If correct, this makes it practical to retain more eye-tracking data in studies of neurodegenerative disease instead of discarding incomplete trials.

What carries the argument

The load-bearing object is the SAITS-RAE pipeline. SAITS (Self-Attention-based Imputation for Time Series) is a transformer-style imputer with two diagonally masked self-attention blocks, so each time step must infer its value from other time steps rather than from itself, followed by a weighted combination block that fuses the two imputation hypotheses. The RAE is a one-dimensional convolutional autoencoder with skip connections trained on complete SPEM sequences, and it refines the upsampled signal to recover fine temporal detail lost in downsampling. The evaluation machinery is equally important: artificial blinks are inserted using empirical distributions of blink duration, position, and count estimated from real recordings, and metrics are computed only at the artificially missing positions, with separate frequency-domain metrics over the whole signal.

What would settle it

Use a second recording modality that does not drop out during blinks, such as a high-speed camera or a second tracker viewing the same eye, to obtain true eye position during real blinks and track losses; impute those real gaps with the SAITS-RAE pipeline and compare point-by-point against the simultaneously recorded truth. If the error on real gaps is not clearly below PCHIP, SSA, and KNN, the central claim fails.

Watch

Extended reading notes

Core claim

The central claim is that combining the SAITS transformer-style imputer with a refinement autoencoder yields the most accurate reconstruction of blink- and track-loss gaps in smooth pursuit eye movement sequences among the methods compared. The authors report SAITS-RAE as the global best in the time domain (MAE 0.10, RMSE 0.13, similarity 0.84 in the standard scenario) and as the strongest method when a continuous 4-second block is missing (MAE 0.36 vs 0.88 for SSA, 1.34 for KNN, 1.64 for PCHIP), while also producing the lowest errors in low-frequency spectral content. The claim is not that deep learning is always better; the paper explicitly notes that KNN often preserves the overall spectral envelope slightly better, so the argued advantage is a balanced combination of temporal fidelity, spectral preservation, and robustness to long gaps.

Load-bearing premise

The load-bearing premise is that gaps manufactured from the average blink statistics of this dataset behave like real blink and track-loss gaps in new recordings, so accuracy measured on these artificial gaps transfers to genuinely missing clinical data.

Editorial extensions

If this is right

  • Incomplete SPEM recordings no longer have to be discarded: imputed sequences can feed downstream biomarker extraction, increasing the usable sample size in studies of Parkinson's disease and other movement disorders.
  • The advantage over classical methods grows with gap length, so the method is aimed precisely at track-loss scenarios that defeat local interpolation.
  • Because frequency-domain error stays low, spectral analyses of imputed sequences, including low-frequency components below 1 Hz, remain informative.
  • A single model trained across all smooth pursuit tasks transfers across stimulus types without task-specific retraining, simplifying clinical deployment.
  • The same two-stage impute-then-refine design can be carried over to other long biomedical time series such as EEG and ECG, as the paper itself suggests.

Reading between the lines

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

  • Because the empirical blink statistics used to generate test gaps were estimated from the entire dataset, including the held-out test participants, the reported test accuracy is likely optimistic for truly novel recording conditions; re-estimating the distributions on training participants only would be a sharper test.
  • The RAE is trained on complete sequences and applied uniformly, so it may pull imputed regions toward the manifold of typical SPEM shapes; whether this introduces bias for atypical or pathological signals is not addressed in the paper.
  • A natural next experiment the paper does not run is to check whether imputation quality changes downstream diagnostic accuracy, e.g., Parkinson's vs control classification, rather than only point-wise reconstruction error.
  • The deterministic nature of the pipeline means no confidence intervals are produced for imputed values; an attention-based approach could in principle be extended to output uncertainty, which would be valuable for long gaps.
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.

Referee Report

3 major / 5 minor

Summary. The manuscript proposes a missing-data imputation pipeline for smooth pursuit eye movement (SPEM) recordings, combining SAITS self-attention imputation on downsampled signals with cubic upsampling and a convolutional refinement autoencoder (RAE). It evaluates the pipeline on 5,504 sequences from 172 participants, comparing against PCHIP, SSA, and KNN under artificially inserted blink-like gaps and under a single 4 s large-gap condition. Time-domain and frequency-domain metrics are reported. The central claim is that SAITS-RAE significantly improves reconstruction accuracy over the compared methods while preserving spectral content.

Significance. If the central claim is validated, the proposed pipeline could be useful for preparing SPEM recordings for downstream biomarker extraction in Parkinson's disease studies. The paper's strengths are the use of a real clinical dataset, a transparent description of the blink-detection and artificial-missingness procedure, and the inclusion of both time- and frequency-domain metrics. However, the evidence as presented does not support the 'state of the art' language in the abstract: only three classical baselines are compared, no error bars or significance tests are reported, and the artificial test masks are sampled from blink statistics computed on the full dataset including test participants. These issues are fixable with additional experiments, and the large-interval results suggest the method has genuine potential.

major comments (3)
  1. [Section 3.5.1] The empirical distributions used to generate artificial missing values are extracted from 'all sequences in the dataset,' which includes the 35 held-out test participants. Because SAITS receives the missingness mask as an input, the test masks are drawn from the same distribution that the model saw during training, and the baseline methods cannot exploit these global blink statistics. This makes the test scenario in-distribution by construction and can inflate the SAITS-RAE advantage, particularly for the small margins in Table 2 (MAE 0.10 vs 0.11 and RMSE 0.13 vs 0.14 against KNN-RAE). Please recompute the empirical blink statistics using only training-participant data and re-report Tables 1-3, or otherwise demonstrate that the reported margins are insensitive to this choice.
  2. [Tables 1-3] Tables 1-3 report single aggregate metric values with no per-sequence variability, confidence intervals, or paired significance tests. The abstract's word 'significant' is not supported by any statistical procedure. Given that several Table 2 comparisons are extremely close (e.g., MAE 0.10 vs 0.11 for SAITS-RAE vs KNN-RAE), the authors should report the distribution of per-sequence errors across the test sequences and run paired tests (e.g., Wilcoxon signed-rank) for each metric; otherwise the claimed improvements over KNN cannot be distinguished from noise.
  3. [Section 4 and Abstract] The paper compares SAITS-RAE only against PCHIP, SSA, and KNN. The introduction discusses GAIN, diffusion models, and CycleGAN as related work, and the abstract claims superiority over 'other state of the art techniques.' Without at least one recent deep-learning imputation baseline (e.g., BRITS, GAIN, CSDI, or a Transformer imputer) evaluated on the same data, the state-of-the-art claim is unsupported. Please add such baselines or soften the claim accordingly.
minor comments (5)
  1. [Section 3.2] The sentence 'SAITS has demonstrated state-of-the-art performance ... [11]' cites the SSSD paper [11]; the correct reference for SAITS benchmarks is the SAITS paper [26] or an appropriate benchmark study.
  2. [Section 4.1, Table 1] The text states that 'SAITS-D achieves the lowest point-wise errors (MAE = 0.10, RMSE = 0.14)', but KNN-D also has RMSE = 0.14; please clarify the tie.
  3. [Figure 6] The histogram of blink durations lacks clear axis labels and units; please add them.
  4. [Section 3.6.1, Eq. (2)] The MRE formula divides by x_i; for signals with small absolute values, MRE can be numerically unstable. Please state the threshold or method used to handle near-zero values beyond excluding exact zeros.
  5. [Section 4.2] The RAE MSE values (1.63e-3 vs 3.30e-3) are for reconstructing clean original signals, not for imputation errors. Please clarify this in the text so readers do not interpret them as imputation improvements.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: empirical benchmark with an evaluation-leakage caveat, not a reduction to its inputs.

full rationale

This is an empirical benchmarking paper, not a derivation; the central claim is a measured performance comparison. SAITS is applied from the published SAITS model (reference [26]) and trained with an MSE objective on z-score-normalized, artificially corrupted sequences, while the RAE is trained only on complete sequences. The evaluation protocol in Section 3.5.1 samples artificial blink durations, positions, and counts from empirical distributions computed over all sequences in the dataset, including the held-out test participants. This makes the test missingness pattern in-distribution by construction, and the paper reports no confidence intervals or significance tests for the claimed improvements. However, these are external-validity and statistical-reporting limitations, not circular reductions: the imputed amplitude values are not constructed from the missingness statistics, and no fitted parameter is renamed as a prediction. The citation of reference [20] for blink detection is a preprocessing detail, not a load-bearing self-citation chain. Therefore, no circular step can be exhibited from the paper's equations or citations, and the appropriate circularity finding is none.

Assumptions & free parameters 9 free parameters · 4 assumptions · 0 invented entities

The paper is entirely empirical, so there are no derived quantities requiring free parameters. The listed values are training hyperparameters and design choices. The load-bearing assumptions are the representativeness of the artificial missingness model, the recoverability after downsampling, and the independence of observations across sequences from the same participant. The RAE is a new model component but not a postulated entity in the sense of a new force or particle; it is an engineered artifact with no falsifiable handle outside its empirical evaluation.

free parameters (9)
  • SAITS latent dimension = 256
    Hand-chosen architecture hyperparameter; affects model capacity but not a scientific constant.
  • SAITS attention heads = 4
    Hand-chosen architecture hyperparameter.
  • SAITS dropout probability = 0.2
    Hand-chosen regularization hyperparameter.
  • SAITS learning rate = 4e-4
    Adam optimizer setting, hand-chosen.
  • RAE learning rate = 1e-4
    Hand-chosen optimizer setting.
  • RAE weight decay = 1e-5
    Hand-chosen optimizer setting.
  • Downsampling factor = 30
    Chosen for computational feasibility; the pipeline's accuracy depends on this reduction from 15,000 to 500 samples.
  • KNN number of neighbors k = not reported
    The paper compares against KNN but never states k, which is a free parameter of the baseline and affects fairness and reproducibility.
  • Frequency cutoffs for RMSEFLow and RMSEFHigh = 1 Hz and 5 Hz
    Hand-chosen thresholds for frequency-domain metrics.
assumptions (4)
  • domain assumption Artificial missing segments generated from empirical blink duration, position, and count distributions are representative of real blink and track-loss missingness in SPEM.
    Section 3.5.1; all evaluation is on simulated missingness, so transfer to real gaps is assumed. The empirical distributions were estimated on all sequences including the test set.
  • domain assumption Downsampling by a factor of 30 preserves sufficient information for the imputation and refinement stages to recover the original signal.
    Section 3.1; the SAITS model operates on 500-sample sequences, so the 30x reduction determines what information is available. The RAE is expected to restore high-frequency content.
  • domain assumption Sequences from different eyes, tasks, and sessions of the same participant can be treated as independent observations.
    Section 2.1 states this assumption explicitly; within-participant correlation could inflate the effective sample size and underestimate variance.
  • domain assumption Blink-induced missingness follows a Missing Not at Random (MNAR) mechanism.
    Section 3.5.1 classifies blink gaps as MNAR; if real gaps are MCAR or MAR, the simulated evaluation may not match deployment.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Imputation of Missing Data in Smooth Pursuit Eye Movements Using a Self-Attention-based Deep Learning Approach." pith.science (2026). https://pith.science/paper/2SU2DA2R

@misc{pith2026250600545,
  author       = {Pith},
  title        = {Pith review of: Imputation of Missing Data in Smooth Pursuit Eye Movements Using a Self-Attention-based Deep Learning Approach},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2SU2DA2R}},
  note         = {Machine review of arXiv:2506.00545}
}
read the original abstract

Missing data is a relevant issue in time series, especially in biomedical sequences such as those corresponding to smooth pursuit eye movements, which often contain gaps due to eye blinks and track losses, complicating the analysis and extraction of meaningful biomarkers. In this paper, a novel imputation framework is proposed using Self-Attention-based Imputation networks for time series, which leverages the power of deep learning and self-attention mechanisms to impute missing data. We further refine the imputed data using a custom made autoencoder, tailored to represent smooth pursuit eye movement sequences. The proposed approach was implemented using 5,504 sequences from 172 Parkinsonian patients and healthy controls. Results show a significant improvement in the accuracy of reconstructed eye movement sequences with respect to other state of the art techniques, substantially reducing the values for common time domain error metrics such as the mean absolute error, mean relative error, and root mean square error, while also preserving the signal's frequency domain characteristics. Moreover, it demonstrates robustness when large intervals of data are missing. This method offers an alternative solution for robustly handling missing data in time series, enhancing the reliability of smooth pursuit analysis for the screening and monitoring of neurodegenerative disorders.

Figures

Figures reproduced from arXiv: 2506.00545 by the authors.

Figure 1
Figure 1. Illustration of an SPEM sequence (blue) along with the target (dotted black) for participant HG032_V2, [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Visual stimuli used in the twelve SPT conditions. Participants initially fix on a central white dot for 200 ms [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Illustration of an SPEM sequence (in blue) along with the target (in dotted black) for participant HF021, [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (7 more)
Figure 4
Figure 4. Figure 4: Illustration of the four-step pipeline followed for imputing missing data in SPEM Recordings. [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Details of the architecture of the RAE. The RAE architecture comprises an encoder-decoder structure built from multiple one-dimensional convolutional layers, batch normalisation, and ReLU activation functions. Skip connections between corresponding encoder and decoder …
Figure 6
Figure 6. Figure 6: Histogram corresponding to the duration of the missing values (NaN) for all the sequences in the dataset. [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]
Figure 7
Figure 7. Figure 7: Comparison of Original, Upsampled, and RAE-Refined Signal Segment (Samples 4100–4200). [PITH_FULL_IMAGE:figures/full_fig_p011_7.png]
Figure 8
Figure 8. Figure 8: Spectrograms of (a) Original sequence; (b) Upsampled; and (c) RAE-Refined. [PITH_FULL_IMAGE:figures/full_fig_p012_8.png]
Figure 9
Figure 9. Figure 9: Illustration of an SPEM sequence (in blue) along with the target (in dotted black) for participant HG032_V2, [PITH_FULL_IMAGE:figures/full_fig_p014_9.png]
Figure 10
Figure 10. Figure 10: Reconstruction of SPEM Sequences with Large Missing Intervals Using Different Imputation Methods for [PITH_FULL_IMAGE:figures/full_fig_p015_10.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

32 extracted references · 29 canonical work pages

  1. [1]

    A survey on missing data in machine learning

    Tlamelo Emmanuel, Thabiso Maupong, Dimane Mpoeleng, Thabo Semong, Banyatsang Mphago, and Oteng Tabona. A survey on missing data in machine learning. Journal of Big data, 8:1–37, 2021

    work page 2021

  2. [2]

    In-database data imputation

    Massimo Perini and Milos Nikolic. In-database data imputation. Proceedings of the ACM on Management of Data, 2(1):1–27, 2024

    work page 2024

  3. [3]

    A benchmark for data imputation methods

    Sebastian Jäger, Arndt Allhorn, and Felix Bießmann. A benchmark for data imputation methods. Frontiers in big Data, 4:693674, 2021

    work page 2021

  4. [4]

    Data management in machine learning: Challenges, techniques, and systems

    Arun Kumar, Matthias Boehm, and Jun Yang. Data management in machine learning: Challenges, techniques, and systems. In Proceedings of the 2017 ACM International Conference on Management of Data, pages 1717–1722, 2017

    work page 2017

  5. [5]

    Oculomotor function in patients with parkinson’s disease

    MS Corin, Teresita S Elizan, and Morris B Bender. Oculomotor function in patients with parkinson’s disease. Journal of the neurological sciences, 15(3):251–265, 1972

    work page 1972

  6. [6]

    Abnormalities of smooth pursuit in parkinson’s disease: A systematic review

    Karen Frei. Abnormalities of smooth pursuit in parkinson’s disease: A systematic review. Clinical parkinsonism & related disorders, 4:100085, 2021. 15 Imputation of Missing Data in Smooth Pursuit Eye Movements A PREPRINT

    work page 2021

  7. [7]

    The measurement of visual sampling during real-world activity in parkinson’s disease and healthy controls: A structured literature review

    Samuel Stuart, Lisa Alcock, Brook Galna, Sue Lord, and Lynn Rochester. The measurement of visual sampling during real-world activity in parkinson’s disease and healthy controls: A structured literature review. Journal of neuroscience methods, 222:175–188, 2014

    work page 2014

  8. [8]

    Direct and indirect effects of attention and visual function on gait impairment in parkinson’s disease: influence of task and turning

    Samuel Stuart, Brook Galna, Louise S Delicato, Sue Lord, and Lynn Rochester. Direct and indirect effects of attention and visual function on gait impairment in parkinson’s disease: influence of task and turning. European Journal of Neuroscience, 46(1):1703–1716, 2017

    work page 2017

Show all 32 references
  1. [9]

    Eye movement abnormalities in movement disorders

    Vivek Lal and Daniel Truong. Eye movement abnormalities in movement disorders. Clinical Parkinsonism & Related Disorders, 1:54–63, 2019

    2019

  2. [10]

    Detection of parkinson’s disease by analysis of smooth pursuit eye movements and machine learning

    M Bejani, E Luque-Buzo, J Gomez-García, A Burlaka-Petrash, F Grandas, and J Godino-Llorente. Detection of parkinson’s disease by analysis of smooth pursuit eye movements and machine learning. In Movement Disorders, volume 37, pages S184–S185, 2022

    2022

  3. [11]

    Diffusion-based time series imputation and forecasting with structured state space models

    Juan Miguel Lopez Alcaraz and Nils Strodthoff. Diffusion-based time series imputation and forecasting with structured state space models. arXiv preprint arXiv:2208.09399, 2022

    2022 arXiv

  4. [12]

    Diffecg: A versatile probabilistic diffusion model for ecg signals synthesis

    Nour Neifar, Achraf Ben-Hamadou, Afef Mdhaffar, and Mohamed Jmaiel. Diffecg: A versatile probabilistic diffusion model for ecg signals synthesis. In 2024 IEEE/ACIS 22nd International Conference on Software Engineering Research, Management and Applications (SERA), pages 182–188...

    2024

  5. [13]

    Gain: Missing data imputation using generative adversarial nets

    Jinsung Yoon, James Jordon, and Mihaela Schaar. Gain: Missing data imputation using generative adversarial nets. In International conference on machine learning, pages 5689–5698. PMLR, 2018

    2018

  6. [14]

    Generative adversarial networks for ecg generation, translation, imputation and denoising

    Alaina Mahalanabis. Generative adversarial networks for ecg generation, translation, imputation and denoising. Master’s thesis, University of Waterloo, 2022

    2022

  7. [15]

    Contextual imputation with missing sequence of eeg signals using generative adversarial networks

    Woonghee Lee, Jaeyoung Lee, and Younghoon Kim. Contextual imputation with missing sequence of eeg signals using generative adversarial networks. IEEE Access, 9:151753–151765, 2021

    2021

  8. [16]

    Conditional generation of periodic signals with fourier-based decoder

    Jiyoung Lee, Wonjae Kim, Daehoon Gwak, and Edward Choi. Conditional generation of periodic signals with fourier-based decoder. arXiv preprint arXiv:2110.12365, 2021

    2021 arXiv

  9. [17]

    Deep generative models for physiolog- ical signals: A systematic literature review

    Nour Neifar, Afef Mdhaffar, Achraf Ben-Hamadou, and Mohamed Jmaiel. Deep generative models for physiolog- ical signals: A systematic literature review. Artificial Intelligence in Medicine, page 103127, 2025

    2025

  10. [18]

    Baseline wander removal applied to smooth pursuit eye movements from parkinsonian patients

    Mehdi Bejani, Elisa Luque-Buzo, Arsen Burlaka-Petrash, Jorge A Gómez-García, Julián D Arias-Londoño, F Grandas-Pérez, Jesus Grajal, and Juan Ignacio Godino-Llorente. Baseline wander removal applied to smooth pursuit eye movements from parkinsonian patients. IEEE Access, 11:321...

    2023

  11. [19]

    Monotone piecewise cubic interpolation

    Frederick N Fritsch and Ralph E Carlson. Monotone piecewise cubic interpolation. SIAM Journal on Numerical Analysis, 17(2):238–246, 1980

    1980

  12. [20]

    Estimation of the cyclopean eye from binocular smooth pursuit tests

    Elisa Luque-Buzo, Mehdi Bejani, Julián D Arias-Londoño, Jorge A Gómez-García, Francisco Grandas-Pérez, and Juan I Godino-Llorente. Estimation of the cyclopean eye from binocular smooth pursuit tests. IEEE Transactions on Cognitive and Developmental Systems, 2024

    2024

  13. [21]

    Analysis of time series structure: SSA and related techniques

    Nina Golyandina, Vladimir Nekrutkin, and Anatoly A Zhigljavsky. Analysis of time series structure: SSA and related techniques. CRC press, 2001

    2001

  14. [22]

    Reconstruction of pupil dilation signal during eye blinking events

    Francesco Onorati, M Mauri, V Russo, Luca Mainardi, et al. Reconstruction of pupil dilation signal during eye blinking events. In Proceeding of the 7th International Workshop on Biosignal Interpretation, pages 117–120, 2012

    2012

  15. [23]

    Deep learning for multivariate time series imputation: A survey

    Jun Wang, Wenjie Du, Yiyuan Yang, Linglong Qian, Wei Cao, Keli Zhang, Wenjia Wang, Yuxuan Liang, and Qingsong Wen. Deep learning for multivariate time series imputation: A survey. arXiv preprint arXiv:2402.04059, 2024

    2024 arXiv

  16. [24]

    Long-term missing value imputation for time series data using deep neural networks

    Jangho Park, Juliane Müller, Bhavna Arora, Boris Faybishenko, Gilberto Pastorello, Charuleka Varadharajan, Reetik Sahu, and Deborah Agarwal. Long-term missing value imputation for time series data using deep neural networks. Neural Computing and Applications, 35(12):9071–9091, 2023

    2023

  17. [25]

    Mads: Modulated auto-decoding siren for time series imputation

    Tom Bamford, Elizabeth Fons, Yousef El-Laham, and Svitlana Vyetrenko. Mads: Modulated auto-decoding siren for time series imputation. arXiv preprint arXiv:2307.00868, 2023

    2023 arXiv

  18. [26]

    SAITS: Self-attention-based imputation for time series

    Wenjie Du, David Côté, and Yan Liu. SAITS: Self-attention-based imputation for time series. Expert Systems with Applications, 219:119619, 2023

    2023

  19. [27]

    Inference and missing data

    Donald B Rubin. Inference and missing data. Biometrika, 63(3):581–592, 1976

    1976

  20. [28]

    Statistical analysis with missing data

    Roderick JA Little and Donald B Rubin. Statistical analysis with missing data. John Wiley & Sons, 2019. 16 Imputation of Missing Data in Smooth Pursuit Eye Movements A PREPRINT

    2019

  21. [29]

    Junger and A Ponce De Leon

    W.L. Junger and A Ponce De Leon. Imputation of missing data in time series for air pollutants. Atmospheric Environment, 102:96–104, 2015

    2015

  22. [30]

    Deep imputation of missing values in time series health data: A review with benchmarking

    Maksims Kazijevs and Manar D Samad. Deep imputation of missing values in time series health data: A review with benchmarking. Journal of biomedical informatics, 144:104440, 2023

    2023

  23. [31]

    Dynamic graph con- volutional recurrent imputation network for spatiotemporal traffic missing data

    Xiangjie Kong, Wenfeng Zhou, Guojiang Shen, Wenyi Zhang, Nali Liu, and Yao Yang. Dynamic graph con- volutional recurrent imputation network for spatiotemporal traffic missing data. Knowledge-Based Systems, 261:110188, 2023

    2023

  24. [32]

    Dynamic time warping-based imputation for univariate time series data

    Émilie Poisson Caillault, Alain Lefebvre, André Bigand, et al. Dynamic time warping-based imputation for univariate time series data. Pattern Recognition Letters, 139:139–147, 2020. 17

    2020

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.