Pith. sign in

REVIEW 3 major objections 5 minor 27 references

A synthetic dataset of French electric load curves with temperature conditioning

T0 review · 3 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read A conditional latent diffusion model trained on 17,000 French residential smart-meter curves can generate synthetic one-year load curves that are nearly indistinguishable from real data, preserving temperature sensitivity and resisting…

desk verdict A useful dataset contribution with a solid core evaluation, but the temperature-conditioning claim is weakened by a partly circular augmentation and a few overly clean utility numbers. read the letter →

arxiv 2504.14046 v1 pith:SOQX3TAU submitted 2025-04-18 cs.LG cs.AI

Tahar Nabil , Ghislain Agoua , Pierre Cauchois , Anne De Moliner , Benoît Grossin This is my paper · ORCID
classification cs.LGcs.AI
keywords syntheticdataelectricloadcurveslatentdiffusionmodeltimeseriesgenerationtemperatureconditioningsmartmeterprivacyGDPRFrenchresidentialelectricity
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 claims that a conditional latent diffusion model can generate one-year, 30-minute-resolution synthetic French residential electricity load curves that are almost indistinguishable from real smart-meter data while preserving the relationship between outdoor temperature and consumption. This matters because real smart-meter data are personal data under the GDPR, so researchers and grid operators need a realistic, privacy-preserving substitute to study electrification, self-consumption, and demand flexibility. The paper releases a dataset of 10,000 such curves with contracted power, time-of-use plan, and local temperature, and evaluates it on fidelity, utility, and privacy.

What carries the argument

The engine is a two-stage conditional latent diffusion model. A convolutional autoencoder with compression factor 4 maps each one-year load curve, viewed as an image of shape 1 x 365 x 48, into a low-dimensional latent code, with a vector-quantization term regularizing the latent space; a denoising diffusion probabilistic model with a UNet then learns the distribution of these codes. Temperature is injected by patching the temperature series and applying cross-attention between the latent code (query) and the patched temperature (keys and values) in the decoder, while static labels (contracted power, time-of-use plan) are concatenated to the code in the diffusion stage. This split lets the autoencoder be trained once and the diffusion model handle both static and dynamic conditioning.

What would settle it

Compare the released synthetic load curves against real smart-meter records from a French region or customer segment not represented in the training set; if a 1-nearest-neighbor classifier separates the two groups with accuracy well above 0.5, or if a PatchTST model trained on synthetic data shows a materially larger forecasting error on that real data than a model trained on real data, the fidelity and utility claims are contradicted.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that conditioning a latent diffusion model on both static customer attributes and a time-varying exogenous series yields synthetic load curves whose fidelity approaches the real-data ceiling: the discriminative score drops to 0.037 (versus 0.357 for TimeGAN), training a forecaster on synthetic data matches training on real data (MSE 0.190 versus 0.190), and membership-inference attacks barely beat random. The authors read this as evidence that the released synthetic dataset is a viable stand-in for real smart-meter data for energy modeling applications.

Load-bearing premise

The load-bearing premise is that the 17,000 training load curves fairly represent French residential electricity consumption; the paper gives no information on customer selection, demographics, or exact geographic spread, so a biased training set would make the synthetic dataset biased in the same way.

Editorial extensions

If this is right

  • Researchers can train short-term load forecasting and tariff-classification models on synthetic data and expect errors close to training on real data, as the paper's TSTR results show.
  • The released dataset enables studies of temperature-driven electricity demand at the individual level without accessing GDPR-protected smart-meter records.
  • The conditioning mechanism extends to other static customer attributes, so the same pipeline can generate curves for additional contracted powers, tariffs, or customer segments.
  • The paper's evaluation protocol (fidelity metrics, TSTR, membership-inference attacks) offers a concrete benchmark template for future synthetic time-series releases.

Reading between the lines

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

  • A user should not treat the 10,000 released curves as representative of all French households: the training sample's selection and demographics are undocumented, so any bias in those 17,000 customers propagates into the synthetic set.
  • The architecture's split—exogenous conditioning handled by the decoder, label conditioning by the diffusion UNet—looks reusable: the same trained autoencoder could be paired with a different exogenous series, such as solar irradiance or electricity price, without retraining the whole pipeline. The paper does not demonstrate this.
  • The privacy tests rule out obvious memorization but not formal privacy: without a differential-privacy guarantee, a stronger adversary could still extract training information.
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

3 major / 5 minor

Summary. The paper introduces a synthetic dataset of 10,000 one-year French residential electric load curves at 30-minute resolution, generated by a latent diffusion model conditioned on static labels (contracted power, time-of-use plan) and on an exogenous outdoor temperature series. The authors train on 17,000 real smart-meter load curves, compare against TimeGAN on a held-out test set, and evaluate fidelity (discriminative score, Context-FID, correlation score, thermo-sensitivity gradients), utility (TSTR forecasting with PatchTST, time-of-use classification), and privacy (membership inference attacks, three-sample MMD test, nearest-neighbor distance ratio). They report that LDM outperforms TimeGAN on fidelity, matches TRTR performance in forecasting, and achieves privacy-attack scores close to random, and they release the dataset via Zenodo.

Significance. If the claims hold, the released dataset is a potentially useful resource for energy modeling, since it provides full-year sub-hourly synthetic load curves with static and temperature conditioning, going beyond earlier synthetic smart-meter datasets that are limited to daily profiles or lack temperature covariates. The paper's strengths include a public dataset with DOI, a systematic comparison against a standard baseline on a real hold-out test set, and a multi-pronged privacy evaluation. The main weakness is that the temperature-conditioning evaluation is partly circular, which matters because temperature conditioning is the paper's principal differentiator relative to prior work. The utility results also lack confidence intervals, leaving the strength of the forecasting and classification claims unclear.

major comments (3)
  1. [Section 3.1, Appendices B.1 and D.1.2] The thermo-sensitivity evaluation is circular and therefore does not support the claim that the latent diffusion model 'learned a meaningful conditioning by temperature.' The autoencoder is trained with a data augmentation (Appendix B.1) that shifts the temperature by a random global offset delta and modifies the load by adding -g*delta, with g a random thermo-sensitivity gradient. The evaluation then measures exactly this imposed linear relationship: the gradient metric in Appendix A.3 regresses weekly load deltas on degree-day deltas, and the qualitative experiment in Appendix D.1.2 applies a global -6.25C offset, which has the same form as the augmentation. Unless the distribution of g is disclosed and shown not to be calibrated to the real gradient distribution, the reported match between synthetic and real gradients and the realistic offset response are direct consequences of the augmentation, not evidence that the full model learned a general temperature dependence. The authors should report the distribution of g, run an ablation without augmentation, or evaluate on a temperature transformation not used in training (e.g., regional temperature profiles or a nonlinear/seasonal offset).
  2. [Tables 2 and 4, Section 3.2] The utility claims rest on point estimates without confidence intervals or repeated-seed variation. In Table 4, the LDM and TRTR MSE values are identical to three decimal places at every horizon (0.204, 0.188, 0.177, 0.192), and the MAE differences are in the third decimal; in Table 2, the classification F1 for LDM (0.564) is slightly below TRTR (0.576), while accuracy is slightly above (0.750 vs. 0.740). Without error bars or a significance test, the statements 'quasi-equivalent' to TRTR and 'outperforms' TimeGAN are not statistically supported. Please report means and standard deviations over multiple runs, and where possible a paired test.
  3. [Appendix A.2, Section 4] The paper does not describe how the 17,000 training customers were selected, their geographic distribution across the 94 departments, or their demographic and behavioral characteristics. Since the released synthetic dataset is offered for energy modeling applications, the claim that it represents French residential electricity consumption is not supported by the information provided. Please add the sampling design, a demographic/geographic breakdown, or a comparison of key statistics with national residential statistics; alternatively, explicitly limit the claims to the population actually sampled.
minor comments (5)
  1. [Table 3] Several scores are reported as 0.0 (for example, LDM 6 kV A, misc. ToU, D_profile). This is likely a rounding artifact; please report more significant digits so the reader can distinguish genuinely perfect separation from values below 0.005.
  2. [Appendix A.3] The temperature threshold T_thresh is stated to lie in [14.5, 18] degrees C, but the paper does not explain how the threshold is chosen or whether it varies by department. This matters for comparing thermo-sensitivity gradients across regions and should be clarified.
  3. [Figure 7] The reference and offset temperature curves are plotted on the same axes as the load curves, which makes the magnitude of the -6.25C offset hard to read. A separate temperature panel or a clearly labeled secondary axis would improve interpretability.
  4. [References] The bibliography entry for Chai and Chadney (2024) contains a duplicated URL prefix: 'https://www.https://www.climatechange.ai/...'. Please correct this typographical error.
  5. [Section 3.2] The classification task uses the conditioning label as ground truth for synthetic data. The text should note that a kNN classifier trained on synthetic data may benefit from the labels being exactly the conditioning variables, and should describe the 104-dimensional input representation in a bit more detail.

Circularity Check

1 steps flagged · score 6.0 of 10

Temperature-offset evaluation is circular: the autoencoder is trained to map a global temperature offset to a load offset, and that same transformation is then reported as evidence of learned temperature conditioning.

  1. self definitional [Appendix B.1 (autoencoder data augmentation); Section 3.1 Thermo-sensitivity; Appendix D.1.2]
    "We also add data augmentation when training the autoencoder: with probability 0.5, we shift the temperature by a random global offsetδ, and modify consequently the load curve by adding −g×δ, with g > 0 a random thermo-sensitivity gradient."

    The paper states in Section 2 that exogenous conditioning is handled by the decoder only and does not affect the diffusion loss. The autoencoder is thus explicitly trained so that a global temperature offset δ maps to a load shift of −gδ. The thermo-sensitivity evaluation in Appendix D.1.2 applies exactly this operation ('the temperature undergoes a strong −6.25 ◦C offset throughout the year') and the paper cites the resulting distortion, together with gradient histograms, as evidence that 'the latent diffusion model learned a meaningful conditioning by temperature' (Section 3.1). The observed offset response is therefore guaranteed by the augmentation, not discovered by the diffusion model; the qualitative temperature-conditioning claim rests on the injected linear relationship.

full rationale

The main fidelity, utility, and privacy claims are evaluated against a real hold-out test set (discriminative score, Context-FID, TSTR forecasting, MIA, MMD test), so those results are not circular and provide independent support for the dataset's overall quality. The circularity is confined to the thermo-sensitivity evaluation: because the decoder is trained with an augmentation that imposes a load shift −gδ for a global temperature offset δ, the subsequent offset experiment measures the same transformation that was injected into training. The paper offers this as evidence that the model 'learned a meaningful conditioning by temperature,' which is not supported by that experiment. The random gradient g is not stated to be sampled from the real gradient distribution, so the gradient-histogram comparison is not as directly forced, but it is not enough to rescue the offset-based claim. Overall, a partial circularity is present, warranting a score of 6 rather than higher because the central dataset-quality claims rest on independent external benchmarks.

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

The ledger reflects the paper's reliance on an unspecified representative training sample, a linear temperature-response assumption injected through augmentation, and standard deep learning background results. The main free parameters are the augmentation variables and a few conditioning hyperparameters; no new physical entities are introduced.

free parameters (4)
  • Augmentation temperature offset delta = random, distribution unspecified
    In autoencoder training, temperature is shifted by a random global offset delta and load is adjusted by -g*delta with g > 0 a random thermo-sensitivity gradient; the distributions of delta and g are not specified, yet they directly encode the temperature-load relationship the model learns.
  • Thermo-sensitivity gradient g in augmentation = random > 0, distribution unspecified
    g controls the strength of the linear temperature response injected during training; it is hand-chosen and not fitted to the real test distribution, but the evaluation of thermo-sensitivity uses the same linear model.
  • Temperature scaling = 35 C
    Temperature is scaled by 35 C before patching (Appendix B.1); a hand-chosen normalization that affects the conditioning magnitude.
  • Patch length P = 32
    Exogenous temperature is patched with P = 32; a hand-chosen hyperparameter affecting the temporal resolution of conditioning.
assumptions (4)
  • domain assumption The 17,000 training load curves are representative of French residential electricity consumption for the intended use of the dataset.
    The paper states the training data covers 94 departments in metropolitan France (Appendix A.2) but gives no sampling or selection details; the usefulness of the released dataset depends on this representativeness.
  • domain assumption The relationship between temperature and electricity load is approximately linear below a temperature threshold, as assumed in both the data augmentation (B.1) and the thermo-sensitivity evaluation (A.3).
    The model is trained with a linear adjustment -g*delta, and the evaluation computes a linear regression gradient between load deltas and degree-day deltas; nonlinear effects such as cooling or saturation are not modeled.
  • standard math Standard results from latent diffusion (Rombach et al. 2022), DDPM (Ho et al. 2020), and cross-attention (Nie et al. 2023) hold as described.
    The method directly builds on these published results without re-deriving them.
  • domain assumption The evaluation metrics (discriminative score, Context-FID, correlation score, TSTR) are appropriate proxies for the quality of synthetic load curves for energy modeling.
    The paper adopts these metrics from prior work (Yoon et al. 2019; Jeha et al. 2022; Chai et al. 2024) but does not establish that they capture all task-relevant properties such as extreme events or long-range dependencies.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A synthetic dataset of French electric load curves with temperature conditioning." pith.science (2026). https://pith.science/paper/SOQX3TAU

@misc{pith2026250414046,
  author       = {Pith},
  title        = {Pith review of: A synthetic dataset of French electric load curves with temperature conditioning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SOQX3TAU}},
  note         = {Machine review of arXiv:2504.14046}
}
read the original abstract

The undergoing energy transition is causing behavioral changes in electricity use, e.g. with self-consumption of local generation, or flexibility services for demand control. To better understand these changes and the challenges they induce, accessing individual smart meter data is crucial. Yet this is personal data under the European GDPR. A widespread use of such data requires thus to create synthetic realistic and privacy-preserving samples. This paper introduces a new synthetic load curve dataset generated by conditional latent diffusion. We also provide the contracted power, time-of-use plan and local temperature used for generation. Fidelity, utility and privacy of the dataset are thoroughly evaluated, demonstrating its good quality and thereby supporting its interest for energy modeling applications.

Figures

Figures reproduced from arXiv: 2504.14046 by the authors.

Figure 1
Figure 1. (a) t-SNE 2D projection of original (blue) and synthetic data across all categories for [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Latent diffusion (Rombach et al., 2022) with conditioning on exogenous variables. E: image encoder, D: image decoder, Q, K, V : cross-attention network between latent vector z and exogenous time series u. x is the input data (load curve) and xˆ its reconstruction. Static labels are added by cross-attention to the diffusion UNet, or by concatenation to the latent codes z. B MODEL AND BASELINE DETAILS B.1 LATENT DIFFU… view at source ↗
Figure 3
Figure 3. Average one-year load curves of (left) Latent Diffusion and (right) TimeGAN. [PITH_FULL_IMAGE:figures/full_fig_p012_3.png] view at source ↗
Figures from the paper (9 more)
Figure 4
Figure 4. Figure 4: Average weekly profiles of Latent Diffusion (orange) and TimeGAN (green) against test [PITH_FULL_IMAGE:figures/full_fig_p012_4.png]
Figure 5
Figure 5. Figure 5: Average autocorrelation functions of Latent Diffusion (orange) and TimeGAN (green) [PITH_FULL_IMAGE:figures/full_fig_p013_5.png]
Figure 6
Figure 6. Figure 6: Thermo-sensitivity gradients computed from load curves aggregated at a daily frequency. [PITH_FULL_IMAGE:figures/full_fig_p013_6.png]
Figure 7
Figure 7. Figure 7: Average one-year synthetic load curves in the class 6 kVA, night ToU, with (orange) or [PITH_FULL_IMAGE:figures/full_fig_p014_7.png]
Figure 8
Figure 8. Figure 8: Individual samples (contracted power 6 kVA). [PITH_FULL_IMAGE:figures/full_fig_p015_8.png]
Figure 9
Figure 9. Figure 9: Individual samples (contracted power 9 kVA). [PITH_FULL_IMAGE:figures/full_fig_p016_9.png]
Figure 10
Figure 10. Figure 10: Individual samples (contracted power 12 kVA). [PITH_FULL_IMAGE:figures/full_fig_p017_10.png]
Figure 11
Figure 11. Figure 11: ROC curves in log scale of the membership inference attacks in the (a) black-box and (b) [PITH_FULL_IMAGE:figures/full_fig_p018_11.png]
Figure 12
Figure 12. Figure 12: Nearest neighbor distance ratio for Latent Diffusion and TimeGAN, computed either on [PITH_FULL_IMAGE:figures/full_fig_p019_12.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

27 extracted references · 17 canonical work pages

  1. [1]

    Akiba, S

    T. Akiba, S. Sano, T. Yanase, T. Ohta, and M. Koyama. Optuna: A next-generation hyperparameter optimization framework. In Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining , 2019

    2019

  2. [2]

    Bounliphone, E

    W. Bounliphone, E. Belilovsky, M. B. Blaschko, I. Antonoglou, and A. Gretton. A test of relative similarity for model selection in generative models. In The Fourth International Conference on Learning Representations, 2016

    work page 2016

  3. [3]

    Carlini, S

    N. Carlini, S. Chien, M. Nasr, S. Song, A. Terzis, and F. Tramèr. Membership inference attacks from first principles. In 2022 IEEE Symposium on Security and Privacy (SP), pages 1897--1914, 2022. doi:10.1109/SP46214.2022.9833649

    arXiv 2022

  4. [4]

    Chai and G

    S. Chai and G. Chadney. Faraday: Synthetic smart meter generator for the smart grid. In ICLR 2024 Workshop on Tackling Climate Change with Machine Learning, 2024. URL https://www.https://www.climatechange.ai/papers/iclr2024/43

    work page 2024

  5. [5]

    S. Chai, G. Chadney, C. Avery, P. Grunewald, P. Van Hentenryck, and P. L. Donti. Defining 'good': Evaluation framework for synthetic smart meter data. arXiv preprint arXiv:2407.11785 https://arxiv.org/abs/2407.11785 , 2024

    arXiv 2024

  6. [6]

    Z. Chen, A. M. Amani, X. Yu, and M. Jalili. Control and optimisation of power grids using smart meter data: A review. Sensors, 23 0 (4), 2023. ISSN 1424-8220. doi:10.3390/s23042118

    doi:10.3390/s23042118 2023

  7. [7]

    K. Cho, B. van Merri \"e nboer, C. Gulcehre, D. Bahdanau, F. Bougares, H. Schwenk, and Y. Bengio. Learning phrase representations using RNN encoder -- decoder for statistical machine translation. In Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing ( EMNLP ) , pages 1724--1734, 2014. doi:10.3115/v1/D14-1179

    doi:10.3115/v1/d14-1179 2014

  8. [8]

    Emami, A

    P. Emami, A. Sahu, and P. Graf. Buildingsbench: A large-scale dataset of 900k buildings and benchmark for short-term load forecasting. In Thirty-seventh Conference on Neural Information Processing Systems Datasets and Benchmarks Track, 2023. URL https://openreview.net/forum?id=c5rqd6PZn6

    work page 2023

Show all 27 references
  1. [9]

    Données personnelles, 2024

    Enedis. Données personnelles, 2024. https://www.enedis.fr/donnees-personnelles [In French, accessed on January 14, 2025]

    2024

  2. [10]

    Franceschi, A

    J.-Y. Franceschi, A. Dieuleveut, and M. Jaggi. Unsupervised Scalable Representation Learning for Multivariate Time Series . In Advances in Neural Information Processing Systems, volume 33, pages 4650--4661, 2019. URL https://proceedings.neurips.cc/paper/2019/file/53c6de78244e9...

    2019

  3. [11]

    Hayes, L

    J. Hayes, L. Melis, G. Danezis, and E. De Cristofaro. LOGAN: Membership inference attacks against generative models . arXiv preprint arXiv:1705.07663 https://arxiv.org/abs/1705.07663 , 2017

    2017 arXiv

  4. [12]

    J. Ho, A. Jain, and P. Abbeel. Denoising diffusion probabilistic models. In H. Larochelle, M. Ranzato, R. Hadsell, M. Balcan, and H. Lin, editors, Advances in Neural Information Processing Systems, pages 6840--6851. Curran Associates, Inc., 2020

    2020

  5. [13]

    Global energy and climate model, iea, paris, 2024

    IEA. Global energy and climate model, iea, paris, 2024. https://www.iea.org/reports/global-energy-and-climate-model [Accessed on January 17, 2025]

    2024

  6. [14]

    P. Jeha, M. Bohlke-Schneider, P. Mercado, S. Kapoor, R. S. Nirwan, V. Flunkert, J. Gasthaus, and T. Januschowski. PSA - GAN : Progressive self attention GAN s for synthetic time series. In International Conference on Learning Representations, 2022. URL https://openreview.net/f...

    2022

  7. [15]

    T. Kim, J. Kim, Y. Tae, C. Park, J.-H. Choi, and J. Choo. Reversible instance normalization for accurate time-series forecasting against distribution shift. In International Conference on Learning Representations, 2022. URL https://openreview.net/forum?id=cGDAkQo1C0p

    2022

  8. [16]

    Liang and H

    X. Liang and H. Wang. Synthesis of realistic load data: Adversarial networks for learning and generating residential load patterns. In NeurIPS 2022 Workshop on Tackling Climate Change with Machine Learning, 2022. URL https://www.climatechange.ai/papers/neurips2022/93

    2022

  9. [17]

    Loshchilov and F

    I. Loshchilov and F. Hutter. Decoupled Weight Decay Regularization . In International Conference on Learning Representations, 2019. URL https://openreview.net/forum?id=Bkg6RiCqY7

    2019

  10. [18]

    H. Ni, L. Szpruch, M. Sabate-Vidales, B. Xiao, M. Wiese, and S. Liao. Sig-Wasserstein GANs for time series generation . In Proceedings of the Second ACM International Conference on AI in Finance, ICAIF '21, 2022. doi:10.1145/3490354.3494393

    2022

  11. [19]

    Y. Nie, N. H. Nguyen, P. Sinthong, and J. Kalagnanam. A time series is worth 64 words: Long-term forecasting with transformers. In The Eleventh International Conference on Learning Representations, 2023. URL https://openreview.net/forum?id=Jbdc0vTOcol

    2023

  12. [20]

    Rombach, A

    R. Rombach, A. Blattmann, D. Lorenz, P. Esser, and B. Ommer. High-resolution image synthesis with latent diffusion models. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 10684--10695, 2022

    2022

  13. [21]

    Thorve, Y

    S. Thorve, Y. Y. Baek, S. Swarup, H. Mortveit, A. Marathe, A. Vullikanti, and M. Marathe. High resolution synthetic residential energy use profiles for the United States . Scientific Data, 10 0 (1): 0 76, 2023. ISSN 2052-4463. doi:10.1038/s41597-022-01914-1

    2023 doi

  14. [22]

    Trindade

    A. Trindade. ElectricityLoadDiagrams20112014 . UCI Machine Learning Repository, 2015. DOI : https://doi.org/10.24432/C58C86

    2015 doi

  15. [23]

    van den Oord, O

    A. van den Oord, O. Vinyals, and k. kavukcuoglu. Neural discrete representation learning. In I. Guyon, U. V. Luxburg, S. Bengio, H. Wallach, R. Fergus, S. Vishwanathan, and R. Garnett, editors, Advances in Neural Information Processing Systems, volume 30. Curran Associates, Inc., 2017

    2017

  16. [24]

    Y. Wang, H. Wu, J. Dong, G. Qin, H. Zhang, Y. Liu, Y.-Z. Qiu, J. Wang, and M. Long. Timexer: Empowering transformers for time series forecasting with exogenous variables. In The Thirty-eighth Annual Conference on Neural Information Processing Systems, 2024. URL https://openrev...

    2024

  17. [25]

    G. Woo, C. Liu, A. Kumar, C. Xiong, S. Savarese, and D. Sahoo. Unified training of universal time series forecasting transformers. In Forty-first International Conference on Machine Learning, 2024. URL https://openreview.net/forum?id=Yd8eHMY1wz

    2024

  18. [26]

    J. Yoon, D. Jarrett, and M. Van der Schaar. Time-series generative adversarial networks. In Advances in Neural Information Processing Systems, volume 32, 2019. URL https://proceedings.neurips.cc/paper_files/paper/2019/hash/c9efe5f26cd17ba6216bbe2a7d26d490-Abstract.html

    2019

  19. [27]

    R. Yuan, S. A. Pourmousavi, W. L. Soong, A. J. Black, J. A. R. Liisberg, and J. Lemos-Vinasco. A synthetic dataset of Danish residential electricity prosumers. Scientific Data, 10 0 (1): 0 371, 2023. ISSN 2052-4463. doi:10.1038/s41597-023-02271-3

    2023 doi

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.