Pith. sign in

REVIEW 4 major objections 7 minor 43 references

Is Precise Recovery Necessary? A Task-Oriented Imputation Approach for Time Series Forecasting on Variable Subset

T0 review · 4 major / 7 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read The paper argues that for forecasting on a partial variable set, imputation should be optimized to help the forecast, not to reconstruct the missing data accurately, and presents a joint-learning framework that does so.

desk verdict The core claim is confounded: TOI-VSF changes both the imputation method and the forecaster's training distribution, so the 15% and Oracle-beating numbers aren't actually established. read the letter →

arxiv 2411.09928 v1 pith:XBRAQONS submitted 2024-11-15 cs.LG

Qi Hao , Runchang Liang , Yue Gao , Hao Dong , Wei Fan , Lu Jiang , Pengyang Wang This is my paper · ORCID
classification cs.LG
keywords variablesubsetforecastingtask-orientedimputationjointlearningself-supervisedmultivariatetimeseriesmissingvariablesspatio-temporaldatasensorfailure
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

Variable Subset Forecasting arises when a forecasting model trained on N variables is deployed with only S < N variables available, for example after sensor failures. The paper claims that the standard goal of imputation, faithful recovery of missing series, is the wrong objective here, and that an imputation model steered by the downstream forecasting loss yields better forecasts than both no imputation and, in several configurations, than having the complete variable set. The proposed TOI-VSF couples a self-supervised imputation module (patching, time embedding, self-attention, and TCN-based generation) with a forecasting backbone, trained jointly with a weighted sum of imputation and forecasting losses. Across four backbones and five datasets, the authors report average improvements of roughly 15% over imputation baselines and consistent gains over the Oracle setting. If correct, the work reframes missing-data handling in forecasting as task-oriented feature engineering rather than data recovery.

What carries the argument

The load-bearing object is the joint learning loss $L = \alpha L_{\mathrm{IMP}} + \beta L_{\mathrm{FCST}}$ with $\alpha + \beta = 1$, coupling a self-supervised imputation module to a forecasting model. The imputation module takes a randomly masked variable subset, patches each variable series, embeds time, applies shared multi-head self-attention over variables, and generates all N series with a two-block dilated causal TCN with residual connections; its self-supervised objective $L_{\mathrm{IMP}}$ is mean absolute error against the ground-truth complete series. The forecasting model $F_\Theta$ consumes the generated series and contributes $L_{\mathrm{FCST}}$, the mean absolute error of the prediction over the lookback and horizon windows. Because gradients of $L_{\mathrm{FCST}}$ flow back into the imputer, the imputer is pushed to produce variables that the forecaster finds useful, not variables that resemble the original data; the $L_{\mathrm{IMP}}$ term is claimed to keep the generated series temporally coherent. This two-loss balance is what makes the imputation task-oriented.

What would settle it

Train TOI-VSF as described, then evaluate on held-out subsets that are not random draws at k = 15%, such as all sensors in one contiguous road segment missing or a fixed sensor always absent. If the forecasting error on these held-out subsets is no better, or worse, than simply feeding the available subset directly to the forecasting model, the claim that task-oriented imputation generalizes across variable subsets is refuted. A second direct check is to verify whether the imputed series deviate substantially from the ground truth while predictions still improve; if no such deviation is observed, the task-oriented mechanism is not what is driving the gains.

Watch

Extended reading notes

Core claim

The central claim, stated on the paper's own terms, is that precise imputation is unnecessary and sometimes harmful for Variable Subset Forecasting. TOI-VSF replaces recovery-faithful imputation (minimizing error against ground-truth masks) with an imputation module that is trained to generate variables that improve the forecast of the available subset, while a self-supervised reconstruction term keeps the generated series temporally plausible. The framework is model-agnostic: it wraps existing backbones (MTGNN, ASTGCN, MSTGCN, TGCN) and is evaluated with k = 15% available variables at inference. The paper reports that TOI-VSF outperforms the Partial setting by an average of 15% and outperforms the Oracle (full-variable) setting by an average margin over 5%, arguing that joint learning lets the forecaster guide the imputer toward beneficial variables and away from redundant or negatively correlated ones. Evaluation is on five datasets (METR-LA, TRAFFIC, SOLAR, ECG5000, ETTH1) with MAE and RMSE.

Load-bearing premise

The load-bearing premise is that the random masked-subset sampler used during training (k = 15%, 100 random subsets per epoch) adequately represents the space of variable subsets that will appear at deployment; if a real subset pattern falls outside this sampled space, the imputer may generate misleading variables and the claimed robustness fails.

Editorial extensions

If this is right

  • If TOI-VSF is right, sensor-failure resilience in traffic, energy, and health monitoring can be achieved by learning task-oriented imputers instead of trying to reconstruct failed sensor feeds.
  • Because the framework is model-agnostic, any existing multivariate forecaster can be wrapped to handle partial-variable deployment without architectural changes.
  • The reported gains over the Oracle setting imply that on some data the complete variable set contains information that actively hurts forecasting; subsetting plus task-oriented imputation can act as a learned feature-selection mechanism.
  • The joint-learning ablation shows that pre-training the imputer separately and freezing it at inference loses most of the benefit, so imputation and forecasting must be co-trained.
  • The method maintains performance as the available subset size k decreases, suggesting graceful degradation under severe sensor loss.

Reading between the lines

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

  • Beyond the paper: because the imputer generates variables that optimize a downstream loss, the same module could be applied to other partial-input tasks such as anomaly detection or classification, where the generated variables would be shaped by task success rather than reconstruction error.
  • Beyond the paper: the robustness claim is established for random subset sampling at k = 15%; a natural stress test is deploying on correlated failures (for instance, a contiguous block of sensors) and checking whether the improvement persists, since such patterns are less likely to appear in the random sampler.
  • A manuscript-level observation rather than an inference: the abstract states four datasets, but the experimental section and tables report results on five (METR-LA, TRAFFIC, SOLAR, ECG5000, ETTH1), so the four-datasets phrase should be corrected.
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 / 7 minor

Summary. The paper proposes TOI-VSF, a framework for Variable Subset Forecasting (VSF) in which a self-supervised imputation module is trained jointly with a downstream forecasting model. The imputation module reconstructs entire missing variables from a random available subset, and the total loss is a weighted combination of an imputation reconstruction loss and a forecasting loss. The authors claim that task-oriented imputation outperforms both recovery-focused imputation baselines and the Oracle setting where all variables are available at inference, reporting an average improvement of 15% over baselines and over 5% over Oracle. Experiments use four backbones (ASTGCN, MSTGCN, MTGNN, TGCN) and five datasets, with a k=15% available-variable setting in the main results.

Significance. The problem of forecasting when entire variables are missing at inference is practically relevant for sensor networks and IoT, and the paper's core idea—that imputation should be optimized for the downstream task rather than for accurate recovery—is plausible and worth testing. The framework is model-agnostic and the paper provides a clear description of the two-phase training procedure. A genuine positive result is that Table I shows consistent improvements of TOI-VSF over the no-imputation Partial setting across all backbone/dataset combinations, and the joint-learning idea is presented with useful ablations. However, the headline claims about superiority over recovery-based imputation and over the Oracle setting are not currently established because the experimental comparisons vary the forecaster's training distribution at the same time as the imputation loss. If the confound is resolved with controlled experiments, the paper could make a meaningful contribution; as it stands, the evidence is conditional.

major comments (4)
  1. [Sections IV-B, IV-C, and Table III] The central comparison is confounded: TOI-VSF trains the forecaster F on imputed inputs generated by the jointly trained imputer (Section III-B, Eq. 13-15), whereas the Partial setting, the Oracle setting, and the traditional imputation baselines in Figures 10-13 use a forecaster trained on clean complete data and only apply imputation at inference. The 'w/o jl' ablation in Table III also changes both factors at once—it removes the forecasting loss from the imputer and returns the forecaster to clean-data training—so it does not isolate task-oriented imputation. The large gains over baselines and Oracle could therefore come from training the forecaster on randomly masked/imputed inputs (a form of augmentation or distribution matching), not from the imputation being task-oriented. To support the paper's central claim, the authors need a controlled experiment that holds the forecaster training protocol fixed across imputation methods, e.g., train the forecaster on the outputs of each imputation method in the same joint or sequential protocol, and report the resulting errors.
  2. [Abstract and Section IV-F (Figures 10-13)] The abstract's headline claim of 'outperforming baseline methods by 15% on average' is not backed by any numeric comparison table. The comparisons against the imputation baselines (FDW, KNNE, IIM, TRMF, CSDI, SAITS, SS-GAN, MICE) are shown only as heatmaps in Figures 10-13, without numeric values, error bars, or standard deviations. The reader cannot verify the average improvement or its variability across the four backbones and five datasets. The paper should include a table with mean and standard deviation for every baseline, backbone, and dataset, and specify exactly how the 15% figure is computed.
  3. [Section IV-C and Table II] The claim that TOI-VSF 'consistently outperforms the Oracle setting by an average margin of over 5%' is contradicted by Table II. The per-dataset average improvements (across the four backbones) are only 1.59% and 1.71% for TRAFFIC and 3.08% and 1.34% for ETTH1 on MAE and RMSE, respectively. Thus the 'over 5%' margin does not hold for two of the five datasets, and the word 'consistently' is not supported. The claim should be restricted to the specific datasets and metrics where the margin actually holds, or the paper should lower the stated threshold.
  4. [Table I] The conclusion in Section IV-B that TOI-VSF leads to 'consistent enhancement' and large average improvements is weakened by the absence of statistical significance tests. Several entries have overlapping error bars under a normality assumption; for example, on METR-LA with TGCN, the Partial MAE is 9.92±0.75 and the TOI-VSF MAE is 8.46±0.94, and the ETTH1 rows have standard deviations above 1.0 in many cases. Since each of the 10 runs produces paired results (same backbone and dataset), the authors should report paired significance tests (e.g., Wilcoxon signed-rank or paired t-test) or confidence intervals to support the claim that the improvements are not due to run-to-run variance.
minor comments (7)
  1. [Section IV-A] The text says the experiments use 'four real-world datasets' but then lists five datasets, including ETTH1 as item (v); the abstract also says 'four datasets.' This inconsistency should be corrected.
  2. [References] Reference [27] (Hu et al., IJCAI 2024, 'Reconstructing missing variables for multivariate time series forecasting via conditional generative flows') appears to have overlapping authorship with the current paper, and it addresses a closely related problem. The paper should cite and explicitly discuss the relationship and differences between TOI-VSF and that prior work.
  3. [Eq. (14)] Equation (14) uses the notation '{xi_t:t+Q} - {~xi_t:t+Q}' for the forecasting loss, but the prediction variable was introduced as '^xi_t:t+Q' in Eq. (13); the tilde notation is inconsistent and should be replaced with the hat notation used elsewhere.
  4. [Tables I and III] The ASTGCN/ETTH1 joint-learning MAE is reported as 1.87 in Table I but as 1.98 in Table III; these values should be reconciled.
  5. [Section IV-E] The text states that the weight-sensitivity results are 'depicted from Figure 8 to Figure 9,' but the relevant figures are Figures 5-9 across all datasets; the citation should cover all five figures.
  6. [Section IV-A Implementation Details] The training description is ambiguous: it says the subset is randomly sampled 100 times for broad coverage, but also that each epoch takes one randomly sampled variable subset. Clarify whether 100 subsets are sampled per epoch or once for the entire training run.
  7. [Section IV-A and Eq. (12)] There is a typo 'soda baseline' that should read 'SOTA baseline,' and Eq. (12) contains 'groud-truth' instead of 'ground-truth.'

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the framework is an empirical joint-training method evaluated against external baselines; overlapping-author citations are background-only and not load-bearing.

full rationale

TOI-VSF's central claim is empirical. The objective in Eq. (15), L = α·L_IMP + β·L_FCST, combines the self-supervised imputation MAE (Eq. 12) with the forecasting MAE (Eq. 14), and the evaluation measures forecasting MAE/RMSE on held-out test splits. Optimizing a loss and measuring the same error metric on unseen data is standard supervised learning, not a prediction that reduces to its input by construction. The method does not fit any parameter to the test subsets and then rename the fit as a prediction; subset sampling, training/validation/test splits, and 10-run averaging are explicit. The Oracle and Partial settings are external benchmarks, not definitions of the method. The only self-citation concern is that references [2] and [27] include overlapping authors (W. Fan and P. Wang); however, these citations appear as background or related work and are never used to justify the central claim, so they are not load-bearing. The reported 'w/o jl' ablation (Table III) shows a large drop without joint training, which supports the importance of the joint objective rather than constituting a tautology. A separate experimental concern, that baselines and the Oracle train the forecaster on clean data while TOI-VSF trains it on imputed inputs, is a potential confound affecting attribution of the gain to task-oriented imputation; this is a correctness or experimental-design issue, not circular reasoning under the defined criteria.

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

The paper introduces no new physical or conceptual entities beyond the framework itself. The free parameters are the loss weights, the subset fraction, and the unspecified architecture hyperparameters. The axioms are domain assumptions about variable usefulness, subset coverage, and distribution compatibility between imputed inputs and the forecasting model.

free parameters (3)
  • Loss weights alpha and beta = Approximately 0.5 each, exact value not reported
    Equation 15 balances the imputation loss and the forecasting loss. The authors vary alpha in Figures 5-9 and select an intermediate value, but the exact choice is not given in the text.
  • Available subset fraction k = 15% for main experiments, varied from 5% to 45%
    The main tables use S = 15% of N. This is a manually chosen scenario parameter, and the robustness curves show performance depends on it.
  • Imputation model hyperparameters = Not reported
    Patch length, embedding dimension, number of attention heads, TCN kernel size, dilation factor, hidden units, learning rate, optimizer, and batch size are not specified, so exact reproduction is impossible from the text.
assumptions (4)
  • domain assumption Not all variables contribute positively to forecasting; some introduce bias, redundancy, or negative correlation.
    This motivates the entire task-oriented framing in the Introduction and Section IV-C. If false, precise recovery would be the right objective and the Oracle baseline would not be beatable in principle.
  • domain assumption Randomly sampled variable subsets during training adequately cover the space of possible test subsets.
    The method relies on training-time subset sampling, with S = kN and 100 samples per epoch, to generalize to arbitrary inference subsets (Section III-A1 and Implementation Details).
  • domain assumption The forecasting model can be trained on imputed complete series without an invalidating distribution shift.
    Joint learning in Section III-B feeds generated variables into the forecasting model as if they were real data, which assumes the imputed distribution remains usable for forecasting.
  • ad hoc to paper An intermediate alpha/beta ratio, with alpha approximately equal to beta, is a reliable default across backbones and datasets.
    Section IV-E states the optimal balance may vary with the backbone, but the paper does not provide a principled rule, so the chosen balance is an ad hoc tuning decision.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Is Precise Recovery Necessary? A Task-Oriented Imputation Approach for Time Series Forecasting on Variable Subset." pith.science (2026). https://pith.science/paper/XBRAQONS

@misc{pith2026241109928,
  author       = {Pith},
  title        = {Pith review of: Is Precise Recovery Necessary? A Task-Oriented Imputation Approach for Time Series Forecasting on Variable Subset},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XBRAQONS}},
  note         = {Machine review of arXiv:2411.09928}
}
abstract

Variable Subset Forecasting (VSF) refers to a unique scenario in multivariate time series forecasting, where available variables in the inference phase are only a subset of the variables in the training phase. VSF presents significant challenges as the entire time series may be missing, and neither inter- nor intra-variable correlations persist. Such conditions impede the effectiveness of traditional imputation methods, primarily focusing on filling in individual missing data points. Inspired by the principle of feature engineering that not all variables contribute positively to forecasting, we propose Task-Oriented Imputation for VSF (TOI-VSF), a novel framework shifts the focus from accurate data recovery to directly support the downstream forecasting task. TOI-VSF incorporates a self-supervised imputation module, agnostic to the forecasting model, designed to fill in missing variables while preserving the vital characteristics and temporal patterns of time series data. Additionally, we implement a joint learning strategy for imputation and forecasting, ensuring that the imputation process is directly aligned with and beneficial to the forecasting objective. Extensive experiments across four datasets demonstrate the superiority of TOI-VSF, outperforming baseline methods by $15\%$ on average.

Figures

Figures reproduced from arXiv: 2411.09928 by the authors.

Figure 1
Figure 1. Variable Subset Forecast Problem: The left and the right [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. An example of the inference phase on the ECG5000 dataset [1] utilizing the forecasting backbone MTGNN [13]. During [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Framework Overview. Left: The training phase. For multivariate time series {x i t−L:t } N i=1, certain variables N − S are randomly masked to obtain a variable subset {u i t−L:t } S i=1. The subset ΨS is then fed to the self-supervised learning model to obtain a reconstructed time series {˜u i t−L:t } N i=1. Subsequently, the reconstructed time series is used for the forecasting task, ultimately yielding the predict… view at source ↗
Figures from the paper (12 more)
Figure 4
Figure 4. Figure 4: Inference Phase. Different from the training phase, [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]
Figure 5
Figure 5. Figure 5: Different Weights of Self-supervised Learning Module Loss on METR-LA. The horizontal axis is [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Different Weights of Self-supervised Learning Module Loss on TRAFFIC. The horizontal axis is [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: Different Weights of Self-supervised Learning Module Loss on SOLAR. The horizontal axis is [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]
Figure 8
Figure 8. Figure 8: Different Weights of Self-supervised Learning Module Loss on ECG5000. The horizontal axis is [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]
Figure 9
Figure 9. Figure 9: Different Weights of Self-supervised Learning Module Loss on ETTH1. The horizontal axis is [PITH_FULL_IMAGE:figures/full_fig_p008_9.png]
Figure 10
Figure 10. Figure 10: Performance of different imputation methods on VSF, [PITH_FULL_IMAGE:figures/full_fig_p009_10.png]
Figure 14
Figure 14. Figure 14: Performance of different k values on VSF, utilizing the backbone MTGNN on the METR-LA dataset. The lower the results, the better the performance. 5 10 15 20 25 30 35 40 45 k * 100 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 MAE Value KNNE MICE TRMF IIM SSGAN CSDI SAITS FD…
Figure 15
Figure 15. Figure 15: Performance of different k values on VSF, utilizing the backbone MTGNN on the SOLAR dataset. The lower the results, the better the performance. the forecasting outcome—some variables, especially those negatively correlated, can hinder accurate predictions. Optimal mod…
Figure 16
Figure 16. Figure 16: Performance of different k values on VSF, utilizing the backbone MTGNN on the TRAFFIC dataset. The lower the results, the better the performance. 5 10 15 20 25 30 35 40 45 k * 100 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 MAE Value KNNE MICE TRMF IIM SSGAN CSDI SAITS FD…
Figure 17
Figure 17. Figure 17: Performance of different k values on VSF, utilizing the backbone MTGNN on the ECG5000 dataset. The lower the results, the better the performance. 5 10 15 20 25 30 35 40 45 k * 100 1.5 2.0 2.5 3.0 3.5 MAE Value KNNE MICE TRMF IIM SSGAN CSDI SAITS FDW TOI-VSF (a) MAE 5 …
Figure 18
Figure 18. Figure 18: Performance of different k values on VSF, utilizing the backbone MTGNN on the ETTH1 dataset. The lower the results, the better the performance. exhibit slight variations depending on the backbones used in the experiments. F. Comparison with Imputation Methods on VSF (…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 25 canonical work pages

  1. [27]

    Recon- structing missing variables for multivariate time series forecasting via conditional generative flows,

    X. Hu, W. Fan, K. Yi, P. Wang, Y . Xu, Y . Fu, and P. Wang, “Recon- structing missing variables for multivariate time series forecasting via conditional generative flows,” in Proceedings of the 33rd International Joint Conference on Artificial Intelligence (IJCAI) , Jeju, Korea, 2024

    work page 2024

  2. [1]

    Multi-variate time series forecasting on variable subsets,

    J. Chauhan, A. Raghuveer, R. Saket, J. Nandy, and B. Ravindran, “Multi-variate time series forecasting on variable subsets,” in Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining , ser. KDD ’22. New York, NY , USA: Association for Computing Machinery, 2022, p. 76–86. [Online]. Available: https://doi.org/10.1145/3534678.3539394

    arXiv 2022

  3. [2]

    Dish-ts: a general paradigm for alleviating distribution shift in time series forecast- ing,

    W. Fan, P. Wang, D. Wang, D. Wang, Y . Zhou, and Y . Fu, “Dish-ts: a general paradigm for alleviating distribution shift in time series forecast- ing,” in Proceedings of the AAAI conference on artificial intelligence , vol. 37, no. 6, 2023, pp. 7522–7529

    2023

  4. [3]

    Deep learning for multivariate time series imputation: A survey,

    J. Wang, W. Du, W. Cao, K. Zhang, W. Wang, Y . Liang, and Q. Wen, “Deep learning for multivariate time series imputation: A survey,”

    work page

  5. [4]

    The internet of things: An overview,

    K. Rose, S. Eldridge, and L. Chapin, “The internet of things: An overview,”The internet society (ISOC) , vol. 80, no. 15, pp. 1–53, 2015. JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 12

    work page 2015

  6. [5]

    Time-series forecasting with deep learning: a survey,

    B. Lim and S. Zohren, “Time-series forecasting with deep learning: a survey,” Philosophical Transactions of the Royal Society A , vol. 379, no. 2194, p. 20200209, 2021

    2021

  7. [6]

    A comprehensive survey on imputation of missing data in internet of things,

    D. Adhikari, W. Jiang, J. Zhan, Z. He, D. B. Rawat, U. Aickelin, and H. A. Khorshidi, “A comprehensive survey on imputation of missing data in internet of things,” ACM Computing Surveys, vol. 55, no. 7, pp. 1–38, 2022

    work page 2022

  8. [7]

    Turn waste into wealth: On simultaneous clustering and cleaning over dirty data,

    S. Song, C. Li, and X. Zhang, “Turn waste into wealth: On simultaneous clustering and cleaning over dirty data,” in Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2015, pp. 1115–1124

    work page 2015

Show all 43 references
  1. [8]

    Swapping repair for misplaced attribute values,

    Y . Sun, S. Song, C. Wang, and J. Wang, “Swapping repair for misplaced attribute values,” in 2020 IEEE 36th International Conference on Data Engineering (ICDE). IEEE, 2020, pp. 721–732

    2020

  2. [9]

    Nearest neighbor ensemble,

    C. Domeniconi and B. Yan, “Nearest neighbor ensemble,” in Proceed- ings of the 17th International Conference on Pattern Recognition, 2004. ICPR 2004., vol. 1. IEEE, 2004, pp. 228–231

    2004

  3. [10]

    Learning individual models for imputation,

    A. Zhang, S. Song, Y . Sun, and J. Wang, “Learning individual models for imputation,” in 2019 IEEE 35th International Conference on Data Engineering (ICDE). IEEE, 2019, pp. 160–171

    2019

  4. [11]

    Csdi: Conditional score- based diffusion models for probabilistic time series imputation,

    Y . Tashiro, J. Song, Y . Song, and S. Ermon, “Csdi: Conditional score- based diffusion models for probabilistic time series imputation,” Ad- vances in Neural Information Processing Systems , vol. 34, pp. 24 804– 24 816, 2021

    2021

  5. [12]

    Saits: Self-attention-based imputation for time series,

    W. Du, D. C ˆot´e, and Y . Liu, “Saits: Self-attention-based imputation for time series,” Expert Systems with Applications , vol. 219, p. 119619, 2023

    2023

  6. [13]

    Con- necting the dots: Multivariate time series forecasting with graph neural networks,

    Z. Wu, S. Pan, G. Long, J. Jiang, X. Chang, and C. Zhang, “Con- necting the dots: Multivariate time series forecasting with graph neural networks,” in Proceedings of the 26th ACM SIGKDD international conference on knowledge discovery & data mining , 2020, pp. 753–763

    2020

  7. [14]

    Attention is all you need,

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

    2017

  8. [15]

    Beyond efficient transformer for long sequence time-series forecasting., 2021, 35,

    H. Zhou, S. Zhang, J. Peng, S. Zhang, J. Li, H. Xiong, and W. Z. Informer, “Beyond efficient transformer for long sequence time-series forecasting., 2021, 35,” DOI: https://doi. org/10.1609/aaai. v35i12 , vol. 17325, pp. 11 106–11 115

    2021 doi

  9. [16]

    An empirical evaluation of generic convolutional and recurrent networks for sequence modeling,

    S. Bai, J. Z. Kolter, and V . Koltun, “An empirical evaluation of generic convolutional and recurrent networks for sequence modeling,” arXiv preprint arXiv:1803.01271, 2018

    2018 arXiv

  10. [17]

    Informer: Beyond efficient transformer for long sequence time-series forecasting,

    H. Zhou, S. Zhang, J. Peng, S. Zhang, J. Li, H. Xiong, and W. Zhang, “Informer: Beyond efficient transformer for long sequence time-series forecasting,” in The Thirty-Fifth AAAI Conference on Artificial Intelli- gence, AAAI 2021, Virtual Conference , vol. 35, no. 12. AAAI Pres...

    2021

  11. [18]

    Attention based spatial- temporal graph convolutional networks for traffic flow forecasting,

    S. Guo, Y . Lin, N. Feng, C. Song, and H. Wan, “Attention based spatial- temporal graph convolutional networks for traffic flow forecasting,” in Proceedings of the AAAI Conference on Artificial Intelligence , 2019

    2019

  12. [19]

    Tgcn: time domain graph convolutional network for multiple objects tracking,

    J. Zhang, “Tgcn: time domain graph convolutional network for multiple objects tracking,” arXiv preprint arXiv:2101.01861 , 2021

    2021 arXiv

  13. [20]

    Temporal regularized matrix factorization for high-dimensional time series prediction,

    H.-F. Yu, N. Rao, and I. S. Dhillon, “Temporal regularized matrix factorization for high-dimensional time series prediction,” in Advances in Neural Information Processing Systems , 2016

    2016

  14. [21]

    Generative semi-supervised learning for multivariate time series imputation,

    X. Miao, Y . Wu, J. Wang, Y . Gao, X. Mao, and J. Yin, “Generative semi-supervised learning for multivariate time series imputation,” in Proceedings of the AAAI conference on artificial intelligence , vol. 35, no. 10, 2021, pp. 8983–8991

    2021

  15. [22]

    mice: Multivariate impu- tation by chained equations in r,

    S. Van Buuren and K. Groothuis-Oudshoorn, “mice: Multivariate impu- tation by chained equations in r,” Journal of statistical software, vol. 45, pp. 1–67, 2011

    2011

  16. [23]

    Time series data and recent imputation techniques for missing data: A review,

    A. Zainuddin, M. A. Hairuddin, A. I. M. Yassin, Z. I. A. Latiff, and A. Azhar, “Time series data and recent imputation techniques for missing data: A review,” in 2022 International Conference on Green Energy, Computing and Sustainable Technology (GECOST), 2022, pp. 346–350

    2022

  17. [24]

    Imputation of missing data in time series for air pollutants,

    W. Junger and A. P. De Leon, “Imputation of missing data in time series for air pollutants,” Atmospheric Environment, vol. 102, pp. 96– 104, 2015

    2015

  18. [25]

    Time series data imputation: A survey on deep learning approaches,

    C. Fang and C. Wang, “Time series data imputation: A survey on deep learning approaches,” arXiv preprint arXiv:2011.11347 , 2020

    2011 arXiv

  19. [26]

    Missing data imputation of magdas-9’s ground electromagnetism with supervised machine learning and con- ventional statistical analysis models,

    H. Muhammad Asraf, K. A. Nur Dalila, N. Md Tahir, Z. I. Abd Latiff, M. H. Jusoh, and Y . Akimasa, “Missing data imputation of magdas-9’s ground electromagnetism with supervised machine learning and con- ventional statistical analysis models,” Alexandria Engineering Journal , 2...

    2021

  20. [28]

    Spatial- temporal interplay in human mobility: A hierarchical reinforcement learning approach with hypergraph representation,

    Z. Zhang, Y . Xiao, L. Jiang, D. Yang, M. Yin, and P. Wang, “Spatial- temporal interplay in human mobility: A hierarchical reinforcement learning approach with hypergraph representation,” in Proceedings of the AAAI Conference on Artificial Intelligence , vol. 38, no. 8, 2024, ...

    2024

  21. [29]

    Exploring the limits of transfer learning with a unified text-to-text transformer,

    C. Raffel, N. Shazeer, A. Roberts, K. Lee, S. Narang, M. Matena, Y . Zhou, W. Li, and P. J. Liu, “Exploring the limits of transfer learning with a unified text-to-text transformer,” J. Mach. Learn. Res. , vol. 21, no. 1, jan 2020

    2020

  22. [30]

    Prottrans: Toward understanding the language of life through self-supervised learning,

    A. Elnaggar, M. Heinzinger, C. Dallago, G. Rehawi, Y . Wang, L. Jones, T. Gibbs, T. Feher, C. Angerer, M. Steinegger et al., “Prottrans: Toward understanding the language of life through self-supervised learning,” IEEE transactions on pattern analysis and machine intelligence,...

    2021

  23. [31]

    Data2vec: A general framework for self-supervised learning in speech, vision and language,

    A. Baevski, W.-N. Hsu, Q. Xu, A. Babu, J. Gu, and M. Auli, “Data2vec: A general framework for self-supervised learning in speech, vision and language,” in International Conference on Machine Learning . PMLR, 2022, pp. 1298–1312

    2022

  24. [32]

    Unsupervised repre- sentation learning by predicting image rotations,

    S. Gidaris, P. Singh, and N. Komodakis, “Unsupervised repre- sentation learning by predicting image rotations,” arXiv preprint arXiv:1803.07728, 2018

    2018 arXiv

  25. [33]

    Time-contrastive networks: Self-supervised learning from video,

    P. Sermanet, C. Lynch, Y . Chebotar, J. Hsu, E. Jang, S. Schaal, S. Levine, and G. Brain, “Time-contrastive networks: Self-supervised learning from video,” in 2018 IEEE international conference on robotics and automation (ICRA). IEEE, 2018, pp. 1134–1141

    2018

  26. [34]

    Self-supervised learning for videos: A survey,

    M. C. Schiappa, Y . S. Rawat, and M. Shah, “Self-supervised learning for videos: A survey,” ACM Computing Surveys , vol. 55, no. 13s, pp. 1–37, 2023

    2023

  27. [35]

    Self-supervised learning for time series analysis: Taxonomy, progress, and prospects,

    K. Zhang, Q. Wen, C. Zhang, R. Cai, M. Jin, Y . Liu, J. Zhang, Y . Liang, G. Pang, D. Song et al. , “Self-supervised learning for time series analysis: Taxonomy, progress, and prospects,” arXiv preprint arXiv:2306.10125, 2023

    2023 arXiv

  28. [36]

    A transformer-based framework for multivariate time series representation learning,

    G. Zerveas, S. Jayaraman, D. Patel, A. Bhamidipaty, and C. Eickhoff, “A transformer-based framework for multivariate time series representation learning,” in Proceedings of the 27th ACM SIGKDD conference on knowledge discovery & data mining , 2021, pp. 2114–2124

    2021

  29. [37]

    Pre-training enhanced spatial- temporal graph neural network for multivariate time series forecasting,

    Z. Shao, Z. Zhang, F. Wang, and Y . Xu, “Pre-training enhanced spatial- temporal graph neural network for multivariate time series forecasting,” in Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining , 2022, pp. 1567–1577

    2022

  30. [38]

    Time-series generative ad- versarial networks,

    J. Yoon, D. Jarrett, and M. Van der Schaar, “Time-series generative ad- versarial networks,” Advances in neural information processing systems, vol. 32, 2019

    2019

  31. [39]

    Contrastive learning based self-supervised time-series analysis,

    J. P ¨oppelbaum, G. S. Chadha, and A. Schwung, “Contrastive learning based self-supervised time-series analysis,” Applied Soft Computing, vol. 117, p. 108397, 2022

    2022

  32. [40]

    Self-supervised contrastive pre-training for time series via time-frequency consistency,

    X. Zhang, Z. Zhao, T. Tsiligkaridis, and M. Zitnik, “Self-supervised contrastive pre-training for time series via time-frequency consistency,” Advances in Neural Information Processing Systems , vol. 35, pp. 3988– 4003, 2022

    2022

  33. [41]

    Self-supervised learning for time series analysis: Taxonomy, progress, and prospects,

    K. Zhang, Q. Wen, C. Zhang, R. Cai, M. Jin, Y . Liu, J. Y . Zhang, Y . Liang, G. Pang, D. Song et al. , “Self-supervised learning for time series analysis: Taxonomy, progress, and prospects,” IEEE Transactions on Pattern Analysis and Machine Intelligence , 2024

    2024

  34. [42]

    Spatio-temporal self-supervised learning for traffic flow prediction,

    J. Ji, J. Wang, C. Huang, J. Wu, B. Xu, Z. Wu, J. Zhang, and Y . Zheng, “Spatio-temporal self-supervised learning for traffic flow prediction,” in Proceedings of the AAAI conference on artificial intelligence , vol. 37, no. 4, 2023, pp. 4356–4364

    2023

  35. [2024]

    Available: https://arxiv.org/abs/2402.04059

    [Online]. Available: https://arxiv.org/abs/2402.04059

    arXiv

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.