Pith. sign in

REVIEW 5 major objections 5 minor 29 references

A Learnable Multi-views Contrastive Framework with Reconstruction Discrepancy for Medical Time-Series

T0 review · 5 major / 5 minor · reviewed 2026-08-09 · deepseek-v4-flash

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

Pith's one-line read The paper proposes that reconstruction error from an AE-GAN trained on healthy external data can serve as a disease-probability prior, and that a learnable multi-view contrastive framework built on this prior outperforms seven baselines…

desk verdict Plausible new feature, but the experiments as printed are unreliable: duplicate tables, a contradictory ablation, and an unvalidated reconstruction-error mechanism. read the letter →

arxiv 2501.18367 v1 pith:E33SV55M submitted 2025-01-30 cs.LG cs.AI

Yifan Wang , Hongfeng Ai , Ruiqi Li , Maowei Jiang , Cheng Jiang , Chenzhong Li This is my paper · ORCID
classification cs.LGcs.AI
keywords medicaltimeseriescontrastivelearningself-supervisedAE-GANmulti-viewEEGECGdiseasediagnosis
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 proposes that medical time-series diagnosis can be improved by borrowing healthy-signal knowledge from an external dataset and by replacing hand-crafted contrastive pairs with learnable views. It introduces LMCRD, which first trains an AE-GAN on normal patients from another center, then uses the AE-GAN's reconstruction error on each target sample as an added disease-probability feature, and finally trains a multi-head attention encoder with inter-view and intra-view contrastive losses across subject, trial, epoch, and temporal levels. The claim is that this combined pipeline consistently outperforms seven baselines on Alzheimer's disease, myocardial infarction, and Parkinson's disease datasets, including when only 10% of labels are available. A reader should care because label scarcity and cross-center differences are the main practical obstacles to deploying deep learning on EEG and ECG data.

What carries the argument

The load-bearing mechanism is the reconstruction-discrepancy feature: a generator trained on healthy external samples produces small error $E = \mathrm{MSE}(G(\bar{x}_i), \bar{x}_i)$ for healthy target samples and larger error for diseased ones, and that scalar is concatenated to the raw sequence so the contrastive encoder can use abnormality likelihood as an extra channel. On top of that, the LMCF encoder combines a dilated convolutional network $h(x)$ with a multi-head attention network $g(x)$ that creates $V$ learnable views; these views are trained by inter-view loss (different views of the same sample are negative) and intra-view loss (same view and subject are positive, different subjects negative), alongside subject-, trial-, epoch-, and temporal-wise InfoNCE losses, weighted $1:1:1:1:2$. The multi-head attention is what removes the need to hand-define positive and negative pairs.

What would settle it

Take the external healthy dataset and the target dataset from two different centers, train the AE-GAN on the external healthy samples, and plot the reconstruction-error distributions for target healthy versus target diseased samples; if the two distributions are not well separated, or if healthy samples from a third center produce errors as large as the diseased target samples, then the discrepancy is encoding center identity rather than disease. A second check is to shuffle the reconstruction-error feature across target samples or replace it with errors from an untrained AE-GAN and observe whether the reported accuracy drop disappears, which would show the feature itself is not the cause of the gains.

Watch

Extended reading notes

Core claim

On its own terms, the paper claims that a single framework can transfer healthy priors across medical centers and adaptively build multi-view representations without manually designing positive and negative pairs. The central assertion is that reconstruction error from an AE-GAN trained only on external healthy samples acts as a usable indicator of disease probability in target data, and that integrating this discrepancy into a learnable multi-view contrastive objective yields state-of-the-art accuracy, precision, recall, F1, AUROC, and AUPRC on Alzheimer's EEG, myocardial-infarction ECG, and Parkinson's EEG benchmarks. The paper further claims that the gains persist and even strengthen in an extreme 10% labeled-data regime, and its ablations attribute the improvement to both the AE-GAN knowledge-transfer module and the multi-view contrastive module.

Load-bearing premise

The load-bearing premise is that an AE-GAN trained on healthy people at an external center will reconstruct healthy target patients well and sick target patients poorly, meaning the cross-center recording differences are smaller than the disease-related differences in the signal.

Editorial extensions

If this is right

  • On all three diseases, the full LMCRD model beats TS2Vec, TF-C, Mixing-up, TS-TCC, SimCLR, CLOCS, and COMET across the reported metrics.
  • With only 10% of labels for fine-tuning, LMCRD still leads, for example 94.67% accuracy on AD versus 91.43% for the best baseline COMET.
  • Ablations show that removing the AE-GAN knowledge transfer (LMCRD-LMC) or the multi-view contrastive module (LMCRD-RD) lowers accuracy and F1, so both components contribute.
  • The method uses the full subject-trial-epoch-temporal hierarchy plus an additional view level, unlike baselines that use only a subset of these levels.
  • The reconstruction-discrepancy feature is produced without any labels on the target set, so the transfer step needs only unlabeled target data.

Reading between the lines

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

  • A natural extension is to test the same reconstruction-discrepancy prior on other label-scarce physiological signals, such as sleep staging, seizure detection, or ICU monitoring, where an external healthy cohort is available.
  • The method's success likely depends on the external and target datasets sharing recording hardware and montage; applying it to a target center with very different equipment may require re-estimating the AE-GAN on matched healthy subjects.
  • One could isolate the source of gain more sharply by comparing against an AE-GAN trained on target healthy samples only, which would reveal whether the benefit comes from external transfer or from any accurate normal-distribution model.
  • Because the view contrastive losses are label-free, the framework could be stacked on top of other pretraining objectives, potentially turning it into a general-purpose medical time-series encoder.
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

5 major / 5 minor

Summary. The paper proposes LMCRD, a two-stage framework for medical time-series diagnosis. In the first stage, an AE-GAN is pretrained on normal samples from an external dataset and used to compute a reconstruction error for target samples, which is appended as an extra feature. In the second stage, a learnable multi-view contrastive framework (LMCF) with multi-head attention is trained using subject-wise, trial-wise, epoch-wise, temporal-wise, inter-view, and intra-view contrastive losses, followed by supervised fine-tuning. Experiments on an Alzheimer's disease EEG dataset, a Parkinson's disease EEG dataset (TDBrain), and a myocardial infarction ECG dataset (PTB) are reported against seven baselines under partial fine-tuning, full fine-tuning with 100% labels, and full fine-tuning with 10% labels.

Significance. If the results were validated, the paper would address a practically important problem: learning from label-limited, single-center medical time series with the help of external normal data. The proposed combination of an external-data-based reconstruction discrepancy and a multi-head attention contrastive framework is a reasonable direction, and the reported gains are large. However, the experimental support in the current manuscript is not trustworthy as presented: Tables 2 and 3 are identical duplicates, the ablation table contains internal contradictions, and the AE-GAN reconstruction-error mechanism that is central to the claimed advantage is never directly validated. These issues make it impossible to assess whether the reported state-of-the-art performance is real or a consequence of the extra feature and reporting inconsistencies.

major comments (5)
  1. [§5.2.2, Tables 2 and 3] Tables 2 and 3 are identical duplicates. The paper claims two full fine-tuning results on the AD dataset, but the same table is printed twice. As a result, the full fine-tuning evidence for the central claim that LMCRD consistently outperforms baselines is currently a single table, not two independent sets of results. The authors must identify which table is the correct one, provide verified results for the other, and explain how this duplication occurred.
  2. [Appendix E, Table B3] Table B3 is internally inconsistent and cannot support the stated ablation conclusions. In the FFT 100% block, the LMCRD row shows accuracy 80.48, which is the PFT value from Table 1, not the 93.23 reported in Table 2. In the same block, the LMCRD-0 row shows accuracy 84.50, which would mean the ablated model outperforms the full model. The text in Appendix E states that LMCRD-0 has the lowest accuracy of 77.42 and that the drops for LMCRD-LMC, LMCRD-RD, and LMCRD-0 are 3.06%, 0.88%, and 3.06%, respectively, which does not match the numbers in Table B3. The ablation therefore does not support the claim that the AE-GAN and LMCF components are indispensable.
  3. [§4.1.3, Eqs. (2)-(3)] The reconstruction-error feature is never validated, yet it is a load-bearing part of the method. The paper asserts that because the AE-GAN is pretrained on external healthy samples, the reconstruction error E is smaller for target healthy samples and larger for abnormal target samples. This requires the external-to-target healthy distribution shift to be smaller than the healthy-to-disease signal in the target center. No evidence is supplied: there are no E-by-group statistics, no AUROC of E alone, no comparison of E on external healthy versus target healthy samples, and no sensitivity analysis. Without such evidence, E may encode center identity rather than disease, and appending E to every target sample (Eq. (3)) could inject a center confound into all downstream representations. In addition, baselines receive F features while LMCRD receives F+1 features, so the comparison is confounded by the extra input channel. The authors should provide direct validation of E as a disease indicator and an ablation that controls for the feature-dimension increase.
  4. [§4.2, Eqs. (4), (5), (7), (8), (9)] The contrastive losses as written do not follow the InfoNCE form that the paper claims. In each of these equations, the denominator sums over negative samples only and does not include the positive term from the numerator. For example, Eq. (5) has numerator exp(h1_i · h2_i) but the denominator sums only over j ≠ i, and Eq. (8) similarly omits the positive pair from the partition function. This is a meaningful technical discrepancy: the optimization objective is not standard InfoNCE, and the behavior of the method may differ substantially from the stated design. The authors should correct the formulas to match their implementation and verify whether the implementation also omits the positive terms.
  5. [§5.2, Tables 1-3] The claim that LMCRD consistently outperforms seven baselines in both PFT and FFT settings is not supported by trustworthy experimental evidence as presented. Beyond the duplicated main tables and contradictory ablation, there is no significance test or confidence-interval analysis for the differences between LMCRD and COMET, which is the closest baseline. Given the large standard deviations reported in several rows, e.g., LMCRD accuracy 93.23±5.25 in Table 2, the authors should report paired statistical tests or effect sizes to establish that the improvements are not within noise.
minor comments (5)
  1. [Abstract and §1] The method is referred to as both LMCF and LMCRD in the abstract and introduction; the relationship between the framework name and the overall method name should be clarified and used consistently.
  2. [§4.2, Eq. (9)] The notation in Eq. (9) is under-specified: the meanings of v+, s-, and the summation index j are not clearly defined, and the expression g(x^k_{i,s}) does not show the view argument v even though the output is written as g^k_{i,v,s}.
  3. [Appendix C] The main text mentions both subject-dependent and subject-independent splits for the external AD dataset, while the target dataset description states a patient-independent split; please clarify which split is used for the reported numbers.
  4. [General] The paper states that source code is released at 'xxxxx'; a working repository URL or an explicit statement of availability is needed.
  5. [§4.2] The Subject-wise Contrastive Loss paragraph is duplicated verbatim in the text; remove the duplicate.

Circularity Check

0 steps flagged · score 1.0 of 10

No circular derivation: AE-GAN reconstruction feature is label-free, contrastive losses are standard InfoNCE variants, and the COMET self-citation is an acknowledged design borrowing rather than a load-bearing proof.

full rationale

The derivation chain is self-contained in the relevant sense. The AE-GAN module is pretrained only on healthy samples x'_i from an external dataset D' (Eq. 1 and Sections 4.1.1-4.1.3); the reconstruction error E = MSE(G(bar x_i), bar x_i) on target samples is a scalar feature (Eq. 2) that is concatenated to the input (Eq. 3). No target label or downstream classifier output enters E, and the paper never fits E to target labels before the contrastive stage, so the 'disease probability' feature is not a renamed prediction of the target outcomes. The contrastive losses (Eqs. 4-11) are standard InfoNCE variants over subject/trial/epoch/temporal/view groupings, with no coefficient fitted to the reported test metrics. The only author self-citation is to COMET [26], used explicitly as a design source for hierarchical levels ('we adopt the concept from COMET'); this is an acknowledged borrowing, not a load-bearing proof, and the central SOTA claim rests on external baselines and benchmark tables rather than on the truth of COMET. The duplicated Tables 2 and 3 and the inconsistent ablation row in Table B3 (FFT 100% LMCRD repeating the PFT value 80.48 instead of the 93.23 in Table 2) are serious experimental-reporting problems and should be treated as correctness risks, but they do not constitute a derivation step that reduces to its own inputs. No circular step can be exhibited.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The framework rests on a small set of hand-chosen contrastive hyperparameters (loss weights, view count, temperature, masking ratio) and on the domain assumption that external healthy data can define normalcy for target data. No new physical entities are introduced. The most important uncharged input is the reconstruction-error feature E, a one-dimensional summary of anomaly that the paper treats as a disease probability.

free parameters (5)
  • Loss weights lambda_S:lambda_R:lambda_E:lambda_T:lambda_V = 1:1:1:1:2
    Chosen by hand in Section 4.2; the final loss Eq. (11) depends on these ratios, and no sensitivity analysis is provided.
  • Number of views V = Unspecified; V=2 used in Eq. (8) example
    The inter-view loss assumes V views and the paper's example sets V=2, but the actual V for the reported experiments is not given.
  • Contrastive temperature tau
    Appears in InfoNCE losses Eqs. (4) and (9); the value is required to reproduce the experiments but is not reported.
  • Data augmentation masking ratio
    Random continuous masking is used for epoch and temporal views; the mask length or ratio is not specified.
  • MHA heads and hidden dimensions C and d
    Encoder outputs h_i in R^{T x C} and view outputs g_i in R^{T x V x d}; C, d, and the number of attention heads are not given.
assumptions (4)
  • domain assumption Reconstruction error from an AE-GAN trained on external healthy samples is smaller for healthy target samples and larger for abnormal target samples.
    Section 4.1.3 states this directly after Eq. (2); it requires that cross-center distribution shift is smaller than the disease-related signal, which is not tested.
  • domain assumption External dataset D' and target dataset D share feature characteristics sufficient for knowledge transfer.
    Section 3 assumes D' is collected for the same medical task as D and that they share similar feature characteristics.
  • domain assumption Hierarchical medical time series organization (subject, trial, epoch, temporal) applies to all three target datasets.
    Adopted from COMET [26], Section 4.2 and Appendix A.1; the PTB heartbeat segmentation uses R-peak timing, which may not align cleanly with the epoch definitions used in the losses.
  • standard math InfoNCE contrastive losses with random masking produce representations useful for downstream binary diagnosis.
    This is a background assumption of the contrastive learning literature; it is not proved here.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Learnable Multi-views Contrastive Framework with Reconstruction Discrepancy for Medical Time-Series." pith.science (2026). https://pith.science/paper/E33SV55M

@misc{pith2026250118367,
  author       = {Pith},
  title        = {Pith review of: A Learnable Multi-views Contrastive Framework with Reconstruction Discrepancy for Medical Time-Series},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/E33SV55M}},
  note         = {Machine review of arXiv:2501.18367}
}
read the original abstract

In medical time series disease diagnosis, two key challenges are identified.First, the high annotation cost of medical data leads to overfitting in models trained on label-limited, single-center datasets. To address this, we propose incorporating external data from related tasks and leveraging AE-GAN to extract prior knowledge,providing valuable references for downstream tasks. Second, many existing studies employ contrastive learning to derive more generalized medical sequence representations for diagnostic tasks, usually relying on manually designed diverse positive and negative sample pairs.However, these approaches are complex, lack generalizability, and fail to adaptively capture disease-specific features across different conditions.To overcome this, we introduce LMCF (Learnable Multi-views Contrastive Framework), a framework that integrates a multi-head attention mechanism and adaptively learns representations from different views through inter-view and intra-view contrastive learning strategies.Additionally, the pre-trained AE-GAN is used to reconstruct discrepancies in the target data as disease probabilities, which are then integrated into the contrastive learning process.Experiments on three target datasets demonstrate that our method consistently outperforms seven other baselines, highlighting its significant impact on healthcare applications such as the diagnosis of myocardial infarction, Alzheimer's disease, and Parkinson's disease.

Figures

Figures reproduced from arXiv: 2501.18367 by the authors.

Figure 1
Figure 1. Training Procedure of LMCRD Let the dataset of medical time series be denoted as D, where each series of data originates from a corresponding subject s and trial r. Each subject s participates in multiple trials r. Due to the lengthy nature of medical sequence data, they are segmented into several shorter sequence fragments. The segmented sequence samples ¯xi ∈ R T ×F are input to the model, with T denoting the sequ… view at source ↗
Figure 2
Figure 2. Mechanism of Multi-View Contrastive Learning [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Visualization of Multi-View Representations on the AD Dataset. The figure compares contrastive [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

29 extracted references · 24 canonical work pages

  1. [1]

    M. F. Anjum, S. Dasgupta, R. Mudumbai, A. Singh, J. F. Cavanagh, and N. S. Narayanan. Linear predictive coding distinguishes spectral eeg features of parkinson’s disease. Parkinsonism & related disorders, 79:79–85, 2020. 9 Table 3: Full fine-tuning results of AD datasets. We use 100% and 10% of labeled data for fine-tuning. * presents our approach. Datase...

    work page 2020

  2. [2]

    Asadi-Aghbolaghi, A

    M. Asadi-Aghbolaghi, A. Darbandsari, A. Zhang, A. Contreras-Sanz, J. Boschman, P. Ahmadvand, M. K¨ obel, D. Farnell, D. G. Huntsman, A. Churg, et al. Learning generalizable ai models for multi-center histopathology image classification. NPJ Precision Oncology, 8(1):151, 2024

    work page 2024

  3. [3]

    Brunekreef, E

    J. Brunekreef, E. Marcus, R. Sheombarsing, J.-J. Sonke, and J. Teuwen. Kandinsky conformal prediction: Efficient calibration of image segmentation algorithms. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 4135–4143, 2024

    work page 2024

  4. [4]

    Chang, Z

    Q. Chang, Z. Yan, M. Zhou, H. Qu, X. He, H. Zhang, L. Baskaran, S. Al’Aref, H. Li, S. Zhang, et al. Mining multi-center heterogeneous medical data with distributed synthetic learning. Nature communications, 14(1):5510, 2023

    work page 2023

  5. [5]

    T. Chen, S. Kornblith, M. Norouzi, and G. Hinton. A simple framework for contrastive learning of visual representations. In International conference on machine learning, pages 1597–1607. PMLR, 2020

    2020

  6. [6]

    Choi and P

    H. Choi and P. Kang. Multi-task self-supervised time-series representation learning. Information Sciences, 671:120654, 2024

    work page 2024

  7. [7]

    Eldele, M

    E. Eldele, M. Ragab, Z. Chen, M. Wu, C. K. Kwoh, X. Li, and C. Guan. Time-series representation learning via temporal and contextual contrasting.arXiv preprint arXiv:2106.14112, 2021

    arXiv 2021

  8. [8]

    Eldele, M

    E. Eldele, M. Ragab, Z. Chen, M. Wu, C.-K. Kwoh, X. Li, and C. Guan. Self-supervised contrastive representation learning for semi-supervised time-series classification. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2023

    work page 2023

Show all 29 references
  1. [9]

    Escudero, D

    J. Escudero, D. Ab´ asolo, R. Hornero, P. Espino, and M. L´ opez. Analysis of electroencephalograms in alzheimer’s disease patients with multiscale entropy.Physiological measurement, 27(11):1091, 2006

    2006

  2. [10]

    Ganin, E

    Y. Ganin, E. Ustinova, H. Ajakan, P. Germain, H. Larochelle, F. Laviolette, M. March, and V. Lempitsky. Domain-adversarial training of neural networks.Journal of machine learning research, 17(59):1–35, 2016

    2016

  3. [11]

    Goldberger, L

    A. Goldberger, L. Amaral, L. Glass, J. Hausdorff, P. C. Ivanov, R. Mark, J. Mietus, G. Moody, C. Peng, H. Stanley, et al. Physionet: Components of a new research resource for complex physiologic signals. circ. 101 (23): e215-e220//circulation electronic pages. 2000. june 13.[Э...

    2000

  4. [12]

    Kiyasseh, T

    D. Kiyasseh, T. Zhu, and D. A. Clifton. Clocs: Contrastive learning of cardiac signals across space, time, and patients. In International Conference on Machine Learning, pages 5606–5615. PMLR, 2021

    2021

  5. [13]

    A. B. L. Larsen, S. K. Sønderby, H. Larochelle, and O. Winther. Autoencoding beyond pixels using a learned similarity metric. InInternational conference on machine learning, pages 1558–1566. PMLR, 2016

    2016

  6. [14]

    P. P. Liang, Z. Deng, M. Q. Ma, J. Y. Zou, L.-P. Morency, and R. Salakhutdinov. Factorized contrastive learning: Going beyond multi-view redundancy.Advances in Neural Information Processing Systems, 36, 2024

    2024

  7. [15]

    Linardos, K

    A. Linardos, K. Kushibar, S. Walsh, P. Gkontra, and K. Lekadir. Federated learning for multi-center imaging diagnostics: a simulation study in cardiovascular disease.Scientific Reports, 12(1):3551, 2022

    2022

  8. [16]

    Liu and S

    J. Liu and S. Chen. Timesurl: Self-supervised contrastive learning for universal time series representation learning. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, pages 13918–13926, 2024

    2024

  9. [17]

    X. Liu, F. Zhang, Z. Hou, L. Mian, Z. Wang, J. Zhang, and J. Tang. Self-supervised learning: Generative or contrastive. IEEE transactions on knowledge and data engineering, 35(1):857–876, 2021

    2021

  10. [18]

    Z. Liu, A. Alavi, M. Li, and X. Zhang. Self-supervised contrastive learning for medical time series: A systematic review. Sensors, 23(9):4221, 2023

    2023

  11. [19]

    McInnes, J

    L. McInnes, J. Healy, and J. Melville. Umap: Uniform manifold approximation and projection for dimension reduction. arXiv preprint arXiv:1802.03426, 2018. 10

    2018 arXiv

  12. [20]

    Miltiadous, K

    A. Miltiadous, K. D. Tzimourta, T. Afrantou, P. Ioannidis, N. Grigoriadis, D. G. Tsalikakis, P. Angelidis, M. G. Tsipouras, E. Glavas, N. Giannakeas, et al. A dataset of eeg recordings from: Alzheimer’s disease, frontotemporal dementia and healthy subjects.OpenNeuro, 1:88, 2023

    2023

  13. [21]

    A. v. d. Oord, Y. Li, and O. Vinyals. Representation learning with contrastive predictive coding. arXiv preprint arXiv:1807.03748, 2018

    2018 arXiv

  14. [22]

    Schneider, C

    T. Schneider, C. Qiu, M. Kloft, D. A. Latif, S. Staab, S. Mandt, and M. Rudolph. Detecting anomalies within time series using local neural transformations.arXiv preprint arXiv:2202.03944, 2022

    2022 arXiv

  15. [23]

    Van Dijk, G

    H. Van Dijk, G. Van Wingen, D. Denys, S. Olbrich, R. Van Ruth, and M. Arns. The two decades brainclinics research archive for insights in neurophysiology (tdbrain) database.Scientific data, 9(1):333, 2022

    2022

  16. [24]

    Wagner, N

    P. Wagner, N. Strodthoff, R.-D. Bousseljot, D. Kreiseler, F. I. Lunze, W. Samek, and T. Schaeffter. Ptb-xl, a large publicly available electrocardiography dataset.Scientific data, 7(1):1–15, 2020

    2020

  17. [25]

    G. Wang, X. Liu, J. Shen, C. Wang, Z. Li, L. Ye, X. Wu, T. Chen, K. Wang, X. Zhang, et al. A deep-learning pipeline for the diagnosis and discrimination of viral, non-viral and covid-19 pneumonia from chest x-ray images. Nature biomedical engineering, 5(6):509–521, 2021

    2021

  18. [26]

    Y. Wang, Y. Han, H. Wang, and X. Zhang. Contrast everything: A hierarchical contrastive framework for medical time-series. Advances in Neural Information Processing Systems, 36, 2024

    2024

  19. [27]

    M. Xu, P. Baraldi, X. Lu, and E. Zio. Generative adversarial networks with adaboost ensemble learning for anomaly detection in high-speed train automatic doors. IEEE Transactions on Intelligent Transportation Systems, 23(12):23408–23421, 2022

    2022

  20. [28]

    Z. Yue, Y. Wang, J. Duan, T. Yang, C. Huang, Y. Tong, and B. Xu. Ts2vec: Towards universal representation of time series. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 36, pages 8980–8987, 2022

    2022

  21. [29]

    Zhang, Q

    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. 11 A Preliminary Kno...

    2024

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.