Pith. sign in

REVIEW 2 major objections 5 minor 62 references

TimeCHEAT: A Channel Harmony Strategy for Irregularly Sampled Multivariate Time Series Analysis

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

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

Pith's one-line read Channel sharing and channel independence do not have to be mutually exclusive: applied locally inside patches and globally across patches, they yield competitive or better results on irregular multivariate time series analysis.

desk verdict A genuinely new local/global channel-strategy split for irregular time series, but the current version is not fully reproducible and the empty-patch failure mode is underexplored. read the letter →

arxiv 2412.12886 v1 pith:2DTDBN2B submitted 2024-12-17 cs.LG

Jiexi Liu , Meng Cao , Songcan Chen This is my paper · ORCID
classification cs.LG
keywords irregularlysampledtimeserieschannel-dependentchannel-independentmultivariatebipartitegraphattentionpatchembeddingtransformer
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 argues that the two standard ways of handling multiple channels in multivariate time series—channel-dependent (CD), which lets channels share information, and channel-independent (CI), which learns each channel separately—do not have to be mutually exclusive. The proposed model, TimeCHEAT, applies CD inside short sub-series patches to learn time embeddings from neighboring channels, then applies CI across patches with a shared Transformer, so each channel keeps its own global attention pattern. This yields competitive or better accuracy than prior state-of-the-art on classification, interpolation, and forecasting of irregularly sampled data, with the key design being the conversion of embedding learning into a bipartite-graph edge-weight prediction problem. The reader should care because the result suggests a task-general recipe for irregular multivariate data that avoids both the failure of pure CI under sampling sparsity and the loss of channel diversity in pure CD models.

What carries the argument

The central mechanism is the channel-harmony split: local channel-dependent embedding via an Irregularity-to-Regularity Graph (I2RGraph) and global channel-independent Transformer encoding. I2RGraph constructs, for each patch, a bipartite graph with channel nodes, observed timestamp nodes, and K reference timestamp nodes; initial observed edges carry a (value, indicator) pair while reference edges carry (0,0), and a multi-head graph attention network updates node and edge features so the learned reference-edge weights form the patch embedding. The CI Transformer then shares weights across channels but processes each channel's patch sequence with its own position encoding, yielding per-channel attention patterns.

What would settle it

Train the model on an ISMTS dataset at several patch lengths P and reference counts K (for example P in {8, 16, 32} and K in {4, 8, 16}) and hold out channels with very low observation density; if performance on sparse channels degrades sharply when the channel-ID embedding is removed, or if the optimal P and K vary so much across datasets that no consistent trend emerges, then the fixed-length learnable reference embedding rather than the local/global split may be carrying the result.

Watch

Extended reading notes

Core claim

TimeCHEAT establishes that channel strategies can be mixed by scale: CD locally and CI globally. Within each patch, a bipartite graph connects channel nodes to observation timestamps and to K fixed reference timestamps; a graph attention network learns the reference edges, producing a fixed-length patch embedding without imposing the common assumption that importance decays with time distance. Across patches, a CI Transformer—shared weights but separate per-channel operation—learns individualized attention patterns. Experiments on P19, P12, PAM, PhysioNet, USHCN, MIMIC-III, and MIMIC-IV report state-of-the-art or competitive results, for example P19 AUROC $89.5 \pm 1.9$ versus ViTST $89.2 \pm 2.0$, PAM accuracy 96.5 versus 95.8, and interpolation MSE at 90% observations $3.686 \times 10^{-3}$ versus mTAND-Full $4.798 \times 10^{-3}$.

Load-bearing premise

The load-bearing premise is that a fixed number of learnable reference timestamps per patch, attended over the actual observations in that patch, captures enough information for any downstream task without a prior on how importance decays with time distance; if a patch is too sparse or too short, those learned reference weights may be driven by channel identities or neighboring channels rather than the data, and the claimed generality would not hold.

Editorial extensions

If this is right

  • If the claim holds, CD and CI are complementary along the patch boundary rather than incompatible alternatives.
  • Fixed-length embeddings for irregular data can be learned without time-decay priors, reducing the need for hand-crafted inductive biases.
  • The same representation pipeline serves classification, forecasting, and interpolation, so task-specific encoders become unnecessary.
  • Channels with sparse observations gain context from correlated channels during embedding yet retain individual attention in encoding, directly addressing the failure of pure CI under sparsity.
  • The patch-based design avoids the blank-area memory expansion of image-based methods, making it more space-efficient for long or highly irregular series.

Reading between the lines

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

  • The local/global split may generalize further than the paper tests: one could apply CI at intermediate granularities or CD at multiple patch sizes, and the argument suggests tuning patch length P and reference count K per dataset rather than fixing them.
  • Because edge weights are learned without a decay prior, the same I2RGraph module could be transplanted to other irregular-input domains such as point clouds or event streams, though the paper does not test this.
  • The reported stability (low standard deviations) hints at lower variance than image-based baselines, but this is only observed on a few datasets; a compute-matched comparison on longer horizons and higher missingness would be needed to confirm the robustness claim.
  • The absence of sensitivity analysis over patch length P and reference point count K leaves open whether the performance gains are robust or pegged to the chosen hyperparameters.
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

2 major / 5 minor

Summary. The paper introduces TimeCHEAT, a transformer-based model for irregularly sampled multivariate time series (ISMTS). The key idea is to split each channel's series into patches, learn a fixed-length representation for each patch using a bipartite graph attention network with channel-dependent (CD) aggregation over observed timestamps and reference points, and then apply a channel-independent (CI) transformer encoder across patches. The authors claim this 'channel harmony' balances cross-channel information and individual channel dynamics. They evaluate on classification (P19, P12, PAM), interpolation (PhysioNet), and forecasting (USHCN, MIMIC-III, MIMIC-IV, Physionet12), reporting five-run means and standard deviations, and they provide an ablation study. The central claim is that the method is competitive with or better than prior state-of-the-art across these tasks.

Significance. If the local-CD/global-CI design is robust, it offers a useful new perspective on channel strategies for ISMTS and a task-general embedding mechanism that avoids explicit exponential-decay assumptions. The paper benefits from a broad evaluation across three tasks, five-run statistics, and a clear ablation showing the role of each component. However, the significance is tempered by the lack of analysis of patch sparsity and hyperparameter sensitivity: at high missingness ratios, the patch embedding can degrade to an instance-independent constant for empty patches, and no evidence is provided that the reported performance is stable to the choice of P and K. The phrase 'without introducing special inductive bias' is also stronger than what the architecture supports.

major comments (2)
  1. [Patch Embedding Learning (Eqs. 3–7)] The robustness of the local CD embedding to empty patches is not demonstrated. If a patch contains no observed timestamps in any channel, then VT is empty, the only edge set is the zero-initialized Eτ, and after L GNN layers the patch embedding Hp depends only on hnode,0_c = FFN(CM(c)) and hnode,0_t = sin(FFN(t)), i.e., only on the channel-ID and reference-point encodings, which are identical for all instances. With missing ratios of 94.9% (P19), 94.2% (MIMIC-III), and 97.8% (MIMIC-IV), some fraction of patches can be expected to be empty, but the paper reports no statistics on patch occupancy and no sensitivity analysis over patch count P or reference-point count K. Without such analysis, the reported results in Tables 1–3 may reflect favorable choices of P and K rather than the proposed local-CD/global-CI harmony, which is the central claim. I ask the authors to provide patch-occupancy statistics on the evaluated datasets and to show how classification/forecasting performance varies with P and K.
  2. [Contributions and Patch Embedding Learning] The claim that the embedding method 'eliminates the need for special prior knowledge' and 'without introducing special inductive bias' is overstated. The patch-based segmentation, the fixed number K of reference points per patch, and the sinusoidal/FFN encoding of timestamps all encode structural priors about locality and temporal regularity. The paper should either provide a sensitivity analysis demonstrating independence from these choices or revise the claim to specify which priors are avoided (e.g., explicit decay-of-importance with time distance).
minor comments (5)
  1. [Table 1] In the P12 AUROC column, TimeCHEAT (84.5) is the second-best method behind ViTST (85.1) and should be underlined per the stated convention; the text also says the gap is 0.5% but the difference is 0.6 percentage points.
  2. [Table 3] On MIMIC-IV and Physionet12, TimeCHEAT's MSE (0.273 and 0.290) is higher than GraFITi's (0.225 and 0.286) by several reported standard deviations; the statement 'consistently ranking within the top two' should be accompanied by a statistical comparison or a softer characterization such as 'competitive'.
  3. [Main Classification Results] The paper claims lower time and space complexity than ViTST, but no runtime or memory measurements are provided; please add such a comparison or qualify the claim.
  4. [Preliminaries] The notation d = N × C × T is confusing because the dataset has N instances of dimension C×T; please clarify that d refers to the flattened size of the whole dataset or rephrase the dimensions.
  5. [Related Work] The related work section cites Raindrop and LIFE but does not explicitly position TimeCHEAT's channel-harmony idea against their graph/feature-selection mechanisms; the 'first to explore channel strategies' claim needs a more precise scope to avoid overclaiming novelty.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; the proposed architecture is evaluated on held-out test data and no reported result reduces to a fitted input by construction.

full rationale

TimeCHEAT's derivation chain is self-contained: it proposes an architecture (patch-level bipartite-graph embedding followed by a per-channel Transformer encoder) and evaluates it on fixed train/validation/test splits against external baselines. No quantity that is claimed as a prediction is defined in terms of the target: the reference-point edge weights Eτ are learned from observed values and task supervision, but they are not fitted to the test labels, and the reported classification, interpolation, and forecasting numbers come from held-out evaluation. The learnable channel-ID matrix CM is a standard learned parameter rather than a label-derived fit, so it does not make the local-CD/global-CI claim true by definition. Citations to GraFITi, mTAND, and other prior work are not self-citations, and none is used to forbid alternatives or to import an unverified uniqueness theorem. The absence of sensitivity analysis over patch length P and reference-point count K, or the possibility of instance-independent empty patches under extreme missingness, is a robustness and completeness limitation, not a circularity: it does not make any reported result equal its input by construction. Therefore no circular step is identified.

Assumptions & free parameters 4 free parameters · 3 assumptions · 1 invented entities

The central claim rests on several architectural choices (patch length, reference point count, graph depth, channel correlation matrix) that are not justified by theory and are not fully disclosed in the arXiv version. The method builds on prior graph-based time series work (GraFITi, You et al.) but the local-global channel split is new. No formal guarantees are provided, and the paper's own ablations indicate that the channel correlation matrix and the CI encoder both contribute to performance, implying the full design is needed for the reported results.

free parameters (4)
  • Patch length P = Not reported in arXiv version (appendix missing)
    The patch segmentation is the core architectural choice; no sensitivity analysis or selected value is given in the main text.
  • Number of reference points K per patch = Not reported
    K defines the embedding dimension and is chosen by the authors; no justification or sweep is provided.
  • Channel correlation matrix CM = Learned during training, initialized as identity
    A learnable parameter in Eq. (3) that encodes channel correlations; it is learned from data and is necessary for the local CD behavior.
  • Number of GNN layers L = Not reported
    Depth of the bipartite graph network; not specified in the main text.
assumptions (3)
  • domain assumption Local temporal smoothness: within a patch, nearby observations are more relevant than distant ones, and aggregating them via a graph improves embeddings.
    This is the motivation for the patch-based CD strategy. The paper itself states that a common assumption (time decay) is avoided, but it replaces it with a localized patch assumption without empirical or theoretical validation.
  • domain assumption The bipartite graph attention network can learn meaningful edge weights to reference points from observed edges, without additional inductive bias.
    The I2RGraph module (Eq. 2 and Fig. 2) is assumed to generalize from observed edges to unobserved reference edges. This is a learned capacity assertion that is not proven and may depend heavily on training data density.
  • ad hoc to paper Channel identities contain useful correlation information that can be learned from an identity-initialized matrix CM.
    The paper claims that one-hot encodings are insufficient and that a learnable CM is better. This is a design assumption that is only tested through the ablation 'w/o correlation' (Table 4), which shows a drop, but it is not theoretically grounded.
invented entities (1)
  • I2RGraph
    purpose: A graph module that learns continuous-time embeddings by predicting edge weights from channel/timestamp nodes to reference points.
    This is a novel module introduced by the paper. It is not a physical entity but a model component. It does not carry independent falsifiable predictions outside the benchmarks, so independent evidence is false (it is only evaluated as part of the whole model).

how reviews work

0 comments
Cite this review

Pith. "Pith review of TimeCHEAT: A Channel Harmony Strategy for Irregularly Sampled Multivariate Time Series Analysis." pith.science (2026). https://pith.science/paper/2DTDBN2B

@misc{pith2026241212886,
  author       = {Pith},
  title        = {Pith review of: TimeCHEAT: A Channel Harmony Strategy for Irregularly Sampled Multivariate Time Series Analysis},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2DTDBN2B}},
  note         = {Machine review of arXiv:2412.12886}
}
read the original abstract

Irregularly sampled multivariate time series (ISMTS) are prevalent in reality. Due to their non-uniform intervals between successive observations and varying sampling rates among series, the channel-independent (CI) strategy, which has been demonstrated more desirable for complete multivariate time series forecasting in recent studies, has failed. This failure can be further attributed to the sampling sparsity, which provides insufficient information for effective CI learning, thereby reducing its capacity. When we resort to the channel-dependent (CD) strategy, even higher capacity cannot mitigate the potential loss of diversity in learning similar embedding patterns across different channels. We find that existing work considers CI and CD strategies to be mutually exclusive, primarily because they apply these strategies to the global channel. However, we hold the view that channel strategies do not necessarily have to be used globally. Instead, by appropriately applying them locally and globally, we can create an opportunity to take full advantage of both strategies. This leads us to introduce the Channel Harmony ISMTS Transformer (TimeCHEAT), which utilizes the CD locally and the CI globally. Specifically, we segment the ISMTS into sub-series level patches. Locally, the CD strategy aggregates information within each patch for time embedding learning, maximizing the use of relevant observations while reducing long-range irrelevant interference. Here, we enhance generality by transforming embedding learning into an edge weight prediction task using bipartite graphs, eliminating the need for special prior knowledge. Globally, the CI strategy is applied across patches, allowing the Transformer to learn individualized attention patterns for each channel. Experimental results indicate our proposed TimeCHEAT demonstrates competitive SOTA performance across three mainstream tasks.

Figures

Figures reproduced from arXiv: 2412.12886 by the authors.

Figure 1
Figure 1. The difference between the 3 kinds of channel [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Overview of TimeCHEAT framework, shown in (a), containing [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

62 extracted references · 47 canonical work pages

  1. [1]

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

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.a...

  2. [2]

    write newline

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

  3. [3]

    K.; Clausel, M.; Horsch, A.; Constant, M.; and Coubez, X

    Agarwal, R.; Sinha, A.; Prasad, D. K.; Clausel, M.; Horsch, A.; Constant, M.; and Coubez, X. 2023. Modelling Irregularly Sampled Time Series Without Imputation. arXiv preprint arXiv:2309.08698

    arXiv 2023

  4. [4]

    M.; Xiao, C.; Zhang, X.; Wang, F.; Jain, A

    Baytas, I. M.; Xiao, C.; Zhang, X.; Wang, F.; Jain, A. K.; and Zhou, J. 2017. Patient subtyping via time-aware LSTM networks. In ACM SIGKDD, 65--74

    work page 2017

  5. [5]

    S.; Januschowski, T.; and G \"u nnemann, S

    Bilo s , M.; Sommer, J.; Rangapuram, S. S.; Januschowski, T.; and G \"u nnemann, S. 2021. Neural flows: Efficient alternative to neural ODEs. NeurIPS, 34: 21325--21337

    work page 2021

  6. [6]

    D.; Hammerschmidt, C

    Camino, R. D.; Hammerschmidt, C. A.; and State, R. 2019. Improving missing data imputation with deep generative models. arXiv preprint arXiv:1902.10666

    arXiv 2019

  7. [7]

    Cao, W.; Wang, D.; Li, J.; Zhou, H.; Li, L.; and Li, Y. 2018. Brits: Bidirectional recurrent imputation for time series. NeurIPS, 31

    work page 2018

  8. [8]

    Che, Z.; Purushotham, S.; Cho, K.; Sontag, D.; and Liu, Y. 2018. Recurrent neural networks for multivariate time series with missing values. Scientific reports, 8(1): 1--12

    work page 2018

Show all 62 references
  1. [9]

    E.; Feng, A.; Hu, W.; Fey, M.; Tassiulas, L.; Leskovec, J.; and Ying, R

    Chen, J.; Lenssen, J. E.; Feng, A.; Hu, W.; Fey, M.; Tassiulas, L.; Leskovec, J.; and Ying, R. 2024. From Similarity to Superiority: Channel Clustering for Time Series Forecasting. arXiv preprint arXiv:2404.01340

    2024 arXiv

  2. [10]

    T.; Rubanova, Y.; Bettencourt, J.; and Duvenaud, D

    Chen, R. T.; Rubanova, Y.; Bettencourt, J.; and Duvenaud, D. K. 2018. Neural ordinary differential equations. NeurIPS, 31

    2018

  3. [11]

    Chen, X.; Zhang, C.; Zhao, X.-L.; Saunier, N.; and Sun, L. 2022. Nonstationary temporal matrix factorization for multivariate time series forecasting. arXiv preprint arXiv:2203.10651

    2022

  4. [12]

    R.; Li, J.; Zhang, X.; Hong, D.; Gupta, R

    Chowdhury, R. R.; Li, J.; Zhang, X.; Hong, D.; Gupta, R. K.; and Shang, J. 2023. Primenet: Pre-training for irregular multivariate time series. In AAAI, volume 37, 7184--7192

    2023

  5. [13]

    De Brouwer, E.; Simm, J.; Arany, A.; and Moreau, Y. 2019. GRU-ODE-Bayes: Continuous modeling of sporadically-observed time series. NeurIPS, 32

    2019

  6. [14]

    Du, W.; C \^o t \'e , D.; and Liu, Y. 2023. Saits: Self-attention-based imputation for time series. Expert Systems with Applications, 219: 119619

    2023

  7. [15]

    Fan, J. 2022. Dynamic Nonlinear Matrix Completion for Time-Varying Data Imputation. In AAAI

    2022

  8. [16]

    L.; Amaral, L

    Goldberger, A. L.; Amaral, L. A.; Glass, L.; Hausdorff, J. M.; Ivanov, P. C.; Mark, R. G.; Mietus, J. E.; Moody, G. B.; Peng, C.-K.; and Stanley, H. E. 2000. PhysioBank, PhysioToolkit, and PhysioNet: components of a new research resource for complex physiologic signals. circul...

    2000

  9. [17]

    Han, L.; Ye, H.-J.; and Zhan, D.-C. 2024. The Capacity and Robustness Trade-off: Revisiting the Channel Independent Strategy for Multivariate Time Series Forecasting. IEEE Transactions on Knowledge and Data Engineering, (01): 1--14

    2024

  10. [18]

    He, K.; Zhang, X.; Ren, S.; and Sun, J. 2016. Deep residual learning for image recognition. In CVPR, 770--778

    2016

  11. [19]

    Horn, M.; Moor, M.; Bock, C.; Rieck, B.; and Borgwardt, K. 2020. Set functions for time series. In ICML, 4353--4363. PMLR

    2020

  12. [20]

    Y.; Lee, J.; Jo, M.; Kook, S.; Jeon, J.; Hyeong, J.; Kim, J.; and Park, N

    Jhin, S. Y.; Lee, J.; Jo, M.; Kook, S.; Jeon, J.; Hyeong, J.; Kim, J.; and Park, N. 2022. Exit: Extrapolation and interpolation-based neural controlled differential equations for time-series classification and forecasting. In ACM Web Conference, 3102--3112

    2022

  13. [21]

    Jin, M.; Zheng, Y.; Li, Y.-F.; Chen, S.; Yang, B.; and Pan, S. 2022. Multivariate time series forecasting with dynamic graph neural odes. IEEE Transactions on Knowledge and Data Engineering

    2022

  14. [22]

    A.; and Mark, R

    Johnson, A.; Bulgarelli, L.; Pollard, T.; Horng, S.; Celi, L. A.; and Mark, R. 2020. Mimic-iv. PhysioNet. Available online at: https://physionet. org/content/mimiciv/1.0/(accessed August 23, 2021), 49--55

    2020

  15. [23]

    J.; Shen, L.; Lehman, L.-w

    Johnson, A.; Pollard, T. J.; Shen, L.; Lehman, L.-w. H.; Feng, M.; Ghassemi, M.; Moody, B.; Szolovits, P.; Celi, L. A.; and Mark, R. G. 2016. MIMIC-III, a freely accessible critical care database Sci. Data, 3(1): 1

    2016

  16. [24]

    Kenton, J. D. M.-W. C.; and Toutanova, L. K. 2019. BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding. In Proceedings of NAACL-HLT, 4171--4186

    2019

  17. [25]

    Kidger, P.; Morrill, J.; Foster, J.; and Lyons, T. 2020. Neural controlled differential equations for irregular time series. NeurIPS, 33: 6696--6707

    2020

  18. [26]

    Le Morvan, M.; Josse, J.; Scornet, E.; and Varoquaux, G. 2021. What’s a good imputation to predict with missing values? NeurIPS, 34: 11530--11540

    2021

  19. [27]

    Li, Z.; Li, S.; and Yan, X. 2023. Time Series as Images: Vision Transformer for Irregularly Sampled Time Series. In NeurIPS

    2023

  20. [28]

    Li, Z.; Qi, S.; Li, Y.; and Xu, Z. 2023. Revisiting long-term time series forecasting: An investigation on linear mapping. arXiv preprint arXiv:2305.10721

    2023 arXiv

  21. [29]

    X.; and Dustdar, S

    Liu, S.; Yu, H.; Liao, C.; Li, J.; Lin, W.; Liu, A. X.; and Dustdar, S. 2021. Pyraformer: Low-complexity pyramidal attention for long-range time series modeling and forecasting. In ICLR

    2021

  22. [30]

    J.; Williams Jr, C.; and Vose, R

    Menne, M. J.; Williams Jr, C.; and Vose, R. S. 2015. United States historical climatology network daily temperature, precipitation, and snow data. Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, Oak Ridge, Tennessee

    2015

  23. [31]

    Montero-Manso, P.; and Hyndman, R. J. 2021. Principles and algorithms for forecasting groups of time series: Locality and globality. International Journal of Forecasting, 37(4): 1632--1653

    2021

  24. [32]

    Murtagh, F.; and Heck, A. 2012. Multivariate data analysis, volume 131. Springer Science & Business Media

    2012

  25. [33]

    Neil, D.; Pfeiffer, M.; and Liu, S.-C. 2016. Phased lstm: Accelerating recurrent network training for long or event-based sequences. NeurIPS, 29

    2016

  26. [34]

    H.; Sinthong, P.; and Kalagnanam, J

    Nie, Y.; Nguyen, N. H.; Sinthong, P.; and Kalagnanam, J. 2023. A Time Series is Worth 64 Words: Long-term Forecasting with Transformers. In ICLR

    2023

  27. [35]

    Reiss, A.; and Stricker, D. 2012. Introducing a new benchmarked dataset for activity monitoring. In 16th international symposium on wearable computers, 108--109. IEEE

    2012

  28. [36]

    A.; Josef, C

    Reyna, M. A.; Josef, C. S.; Jeter, R.; Shashikumar, S. P.; Westover, M. B.; Nemati, S.; Clifford, G. D.; and Sharma, A. 2020. Early prediction of sepsis from clinical data: the PhysioNet/Computing in Cardiology Challenge 2019. Critical care medicine, 48(2): 210--217

    2020

  29. [37]

    T.; and Duvenaud, D

    Rubanova, Y.; Chen, R. T.; and Duvenaud, D. K. 2019. Latent ordinary differential equations for irregularly-sampled time series. NeurIPS, 32

    2019

  30. [38]

    Schirmer, M.; Eltayeb, M.; Lessmann, S.; and Rudolph, M. 2022. Modeling irregular time series with continuous recurrent units. In ICML, 19388--19405. PMLR

    2022

  31. [39]

    N.; and Schmidt-Thieme, L

    Scholz, R.; Born, S.; Duong-Trung, N.; Cruz-Bournazou, M. N.; and Schmidt-Thieme, L. 2022. Latent Linear ODEs with Neural Kalman Filtering for Irregular Time Series Forecasting. NeurIPS

    2022

  32. [40]

    Schulz, M.; and Stattegger, K. 1997. SPECTRUM: Spectral analysis of unevenly spaced paleoclimatic time series. Computers & Geosciences, 23(9): 929--945

    1997

  33. [41]

    N.; and Marlin, B

    Shukla, S. N.; and Marlin, B. 2018. Interpolation-Prediction Networks for Irregularly Sampled Time Series. In ICLR

    2018

  34. [42]

    N.; and Marlin, B

    Shukla, S. N.; and Marlin, B. 2021. Multi-Time Attention Networks for Irregularly Sampled Time Series. In ICLR

    2021

  35. [43]

    N.; and Marlin, B

    Shukla, S. N.; and Marlin, B. 2022. Heteroscedastic Temporal Variational Autoencoder For Irregularly Sampled Time Series. In ICLR

    2022

  36. [44]

    J.; Celi, L

    Silva, I.; Moody, G.; Scott, D. J.; Celi, L. A.; and Mark, R. G. 2012. Predicting in-hospital mortality of icu patients: The physionet/computing in cardiology challenge 2012. In 2012 Computing in Cardiology, 245--248. IEEE

    2012

  37. [45]

    Sun, C.; Li, H.; Song, M.; Cai, D.; Zhang, B.; and Hong, S. 2024. Time pattern reconstruction for classification of irregularly sampled time series. Pattern Recognition, 147: 110075

    2024

  38. [46]

    Tan, C.; Li, S.; Gao, Z.; Guan, W.; Wang, Z.; Liu, Z.; Wu, L.; and Li, S. Z. 2023. Openstl: A comprehensive benchmark of spatio-temporal predictive learning. NeurIPS, 36: 69819--69831

    2023

  39. [47]

    Tang, X.; Yao, H.; Sun, Y.; Aggarwal, C.; Mitra, P.; and Wang, S. 2020. Joint modeling of local and global temporal dynamics for multivariate time series forecasting with missing values. In AAAI, volume 34, 5956--5963

    2020

  40. [48]

    Tashiro, Y.; Song, J.; Song, Y.; and Ermon, S. 2021. CSDI: Conditional Score-based Diffusion Models for Probabilistic Time Series Imputation. NeurIPS, 34

    2021

  41. [49]

    N.; Kaiser, .; and Polosukhin, I

    Vaswani, A.; Shazeer, N.; Parmar, N.; Uszkoreit, J.; Jones, L.; Gomez, A. N.; Kaiser, .; and Polosukhin, I. 2017. Attention is all you need. NeurIPS, 30

    2017

  42. [50]

    Veli c kovi \'c , P.; Cucurull, G.; Casanova, A.; Romero, A.; Lio, P.; and Bengio, Y. 2017. Graph attention networks. arXiv preprint arXiv:1710.10903

    2017 arXiv

  43. [51]

    Wang, J.; Du, W.; Cao, W.; Zhang, K.; Wang, W.; Liang, Y.; and Wen, Q. 2024. Deep Learning for Multivariate Time Series Imputation: A Survey. arXiv preprint arXiv:2402.04059

    2024 arXiv

  44. [52]

    Wu, H.; Xu, J.; Wang, J.; and Long, M. 2021 a . Autoformer: Decomposition transformers with auto-correlation for long-term series forecasting. NeurIPS, 34: 22419--22430

    2021

  45. [53]

    Wu, Y.; Ni, J.; Cheng, W.; Zong, B.; Song, D.; Chen, Z.; Liu, Y.; Zhang, X.; Chen, H.; and Davidson, S. B. 2021 b . Dynamic gaussian mixture based deep generative model for robust forecasting on sparse multivariate time series. In AAAI, volume 35, 651--659

    2021

  46. [54]

    Wu, Z.; Pan, S.; Long, G.; Jiang, J.; Chang, X.; and Zhang, C. 2020. Connecting the dots: Multivariate time series forecasting with graph neural networks. In ACM SIGKDD, 753--763

    2020

  47. [55]

    K.; Madhusudhanan, K.; Scholz, R.; Ahmed, N.; Burchert, J.; Jawed, S.; Born, S.; and Schmidt-Thieme, L

    Yalavarthi, V. K.; Madhusudhanan, K.; Scholz, R.; Ahmed, N.; Burchert, J.; Jawed, S.; Born, S.; and Schmidt-Thieme, L. 2024. GraFITi: Graphs for Forecasting Irregularly Sampled Time Series. In AAAI, 16255--16263

    2024

  48. [56]

    Yoon, J.; Jordon, J.; and Schaar, M. 2018. Gain: Missing data imputation using generative adversarial nets. In ICML, 5689--5698. PMLR

    2018

  49. [57]

    J.; and Leskovec, J

    You, J.; Ma, X.; Ding, Y.; Kochenderfer, M. J.; and Leskovec, J. 2020. Handling missing data with graph representation learning. NeurIPS, 33: 19075--19087

    2020

  50. [58]

    Zeng, A.; Chen, M.; Zhang, L.; and Xu, Q. 2023. Are transformers effective for time series forecasting? In AAAI, volume 37, 11121--11128

    2023

  51. [59]

    Zhang, X.; Zeman, M.; Tsiligkaridis, T.; and Zitnik, M. 2021 a . Graph-Guided Network for Irregularly Sampled Multivariate Time Series. In ICLR

    2021

  52. [60]

    Zhang, Z.-Y.; Zhang, S.-Q.; Jiang, Y.; and Zhou, Z.-H. 2021 b . LIFE: Learning individual features for multivariate time series prediction with missing values. In ICDM, 1511--1516. IEEE

    2021

  53. [61]

    Zhou, H.; Zhang, S.; Peng, J.; Zhang, S.; Li, J.; Xiong, H.; and Zhang, W. 2021. Informer: Beyond efficient transformer for long sequence time-series forecasting. In AAAI, volume 35, 11106--11115

    2021

  54. [62]

    Zhou, T.; Ma, Z.; Wen, Q.; Wang, X.; Sun, L.; and Jin, R. 2022. Fedformer: Frequency enhanced decomposed transformer for long-term series forecasting. In ICML, 27268--27286. PMLR

    2022

Pith tools

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