DLGAN : Time Series Synthesis Based on Dual-Layer Generative Adversarial Networks

Shuhan Liu; Xuan Hou; Yaohui Chu; Yining Wang; Yue Zhang; Zhaohui Peng

REVIEW 3 major objections 4 minor 45 references

DLGAN : Time Series Synthesis Based on Dual-Layer Generative Adversarial Networks

T0 review · 3 major / 4 minor · reviewed 2026-08-05 · deepseek-v4-flash

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

Pith's one-line read DLGAN claims that decomposing time series generation into supervised feature extraction followed by reconstruction lets a GAN synthesize sequences whose temporal dependencies survive, and reports best discriminative and predictive scores ov

desk verdict A sensible two-stage GAN architecture for time series, but the headline numbers are not yet reproducible—worth a careful referee, not a desk reject. read the letter →

arxiv 2508.21340 v1 pith:GMEXT3I5 submitted 2025-08-29 cs.LG cs.AI

Xuan Hou , Shuhan Liu , Zhaohui Peng , Yaohui Chu , Yue Zhang , Yining Wang This is my paper · ORCID

classification cs.LGcs.AI

keywords timeseriessynthesisgenerativeadversarialnetworkstemporalfeatureextractionsequencereconstructionsyntheticdatadependenciessupervisedlearningmultivariate

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

This paper argues that synthetic time series often look wrong because existing GANs, VAEs, and diffusion models start from random sequences and try to learn temporal structure while generating, which weakens temporal dependencies and misses the original series' features. DLGAN is a generative adversarial network built in two stages: a supervised autoencoder first compresses the real time series and reconstructs it, and then a GAN generates feature vectors in that compressed space, with a second GAN reconstructing the hidden sequence from those vectors. The claim is that placing generation in the learned feature space, rather than in raw sequence space, lets the generator inherit temporal dependencies from the original data. If correct, this makes synthetic time series usable as stand-ins for private or commercially sensitive industrial data without sacrificing predictive utility.

What carries the argument

The load-bearing object is the temporal feature bottleneck: a temporal feature extractor that turns a multivariate hidden sequence into a single compact vector by first modeling each channel independently with non-overlapping sliding-window patches and multi-head self-attention plus positional encoding, then modeling cross-channel patch dependencies with self-attention without positional encoding, then passing through a deep GRU. This vector is what the first GAN learns to synthesize and what the second GAN learns to expand back into a hidden sequence. The two-stage decomposition means random noise never touches raw sequence space; the paper argues this is why generated sequences retain temp

What would settle it

Re-run DLGAN and all six baselines under the TimeGAN protocol on the same four datasets across many random seeds; if the discriminative and predictive scores come within noise of one another, or if another baseline ties DLGAN, the central claim of superiority is falsified. A cheaper probe: check whether the supervised reconstruction loss is the active ingredient by training DLGAN without teacher forcing; if temporal fidelity does not drop, the mechanism claimed is not what carries the result.

Watch

Extended reading notes

Core claim

The central discovery DLGAN proposes is that time series synthesis should be decomposed into two stacked generative layers sandwiching a temporal feature bottleneck. The sequence autoencoder maps original series X into hidden sequence H via a deep GRU and reconstructs X, providing supervised signal. The Temporal Feature Generator extracts a compact temporal feature vector from the real hidden sequence using channel-independent patch-based multi-head self-attention and GRU, then trains Generator1 to synthesize feature vectors that fool Discriminator1. The Sequence Reconstructor (Generator2) then autoregressively rebuilds the hidden sequence from either the real or synthesized feature vector,

Load-bearing premise

The comparison assumes the evaluation protocol exactly matches the TimeGAN setup—same preprocessing, train/test splits, hyperparameters, and score computation—so the reported margins reflect model quality rather than setup differences.

Editorial extensions

If this is right

Synthetic time series produced through the feature bottleneck can be shared with outside analysts while preserving temporal dependency structure, addressing the data-circulation use case in industrial settings.
The supervised autoencoder pretraining plus low-dimensional feature-space GAN gives a recipe that should generalize to other sequence domains, not just the four tabular-style datasets tested.
Because the temporal feature extractor operates channel-independently before cross-channel attention, DLGAN scales to high-dimensional multivariate series without losing local temporal dynamics.
Ablation results imply even the simplest sequence modeling process improves synthesis quality when anchored by supervised reconstruction on original data, supporting the paper's core claim that supervision is the main driver.
Reported scores beat six baselines on discriminative and predictive metrics; if replicated, DLGAN becomes the new reference point for time series GAN comparisons.

Reading between the lines

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

The paper's logic suggests an untested extension: applying the same feature-bottleneck decomposition to other generative families (e.g., diffusion or VAE) could improve their temporal fidelity, since the bottleneck, not the GAN itself, may be doing most of the work.
It also implies a testable prediction: synthetic series from DLGAN should show better phase alignment or autocorrelation structure than those from direct-generation GANs, something the discriminative score measures indirectly and could be checked with spectral or autocorrelation statistics.
The reported gains on Stock and Exchange, low-dimensional daily series, are small; whether DLGAN's advantage holds on longer, higher-frequency series is an open question the paper does not address.
Since the paper does not report confidence intervals, a fair comparison across multiple random seeds with the same protocol would reveal whether the observed margins are practically meaningful.

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.

Desk Editor's Note

read the letter

Colleague,

Short version: this is a plausible incremental contribution to synthetic time series generation. The architecture is not deeply novel — sequence autoencoder, a GAN in the embedding space, and a second reconstruction GAN with supervised losses — but the combination is coherent and the paper includes an ablation study that mostly behaves as expected. The problem is that the central claim, that DLGAN beats TimeGAN and others on four datasets, is not currently verifiable from the manuscript.

What the paper does well: it gives a clear rationale for why learning in a latent space first is sensible; the training procedure is specified with loss functions; and the ablation shows that removing the two new components generally degrades discriminative and predictive scores. That is real supporting evidence, even if the margins are small in places.

Where it is soft, and this is the load-bearing issue: Section 5.1 and Table 1. The paper does not report code, data splits, preprocessing, sequence lengths, normalization, hyperparameters, number of seeds, or any variance measure. The baseline numbers look like they were taken from the original papers, but those papers did not all run on ETTH, Exchange, or Weather. If the baselines were not rerun under identical conditions, the comparison is uncontrolled and the reported margins (e.g., discriminative 0.079 vs 0.106 on ETTH) say little. The stress-test note is right: this is an external reproducibility gap, not an internal contradiction.

Other minor issues: the related work is thin and the "first to combine" framing is overclaimed — the actual machinery is a repackaging of known components. Also the generator loss uses supervised reconstruction on the real hidden sequence, which blurs the line between generation and reconstruction; this makes the strong results less surprising but not circular.

Who is this for: researchers working on GAN-based time series generation who want another architecture variant to test. It is not a methods breakthrough, but it is a valid empirical question whether this design helps. I would not cite it until the code and evaluation details are released.

For peer review: yes, send it. The internal logic is fine and the empirical weakness is fixable with code, seeds, confidence intervals, and a controlled baseline rerun. That is the kind of revision a good referee can ask for.

Referee Report

3 major / 4 minor

Summary. The paper proposes DLGAN, a GAN-based time series generator that separates generation into three modules: a sequence autoencoder that maps real series to a hidden space; a temporal feature generator consisting of a feature extractor, Generator1, and Discriminator1 that aligns synthetic feature vectors with real temporal features; and a sequence reconstructor (Generator2/Discriminator2) that autoregressively reconstructs hidden sequences from feature vectors. Training combines supervised MSE reconstruction losses with standard GAN losses, first pretraining the autoencoder and the feature extractor/reconstructor, then jointly training all modules with the combined generator objective in Eq. (14). Experiments on ETTH, Stock, Exchange, and Weather compare DLGAN with six baselines using discriminative and predictive scores adopted from TimeGAN, plus t-SNE visualizations and an ablation study of the two main components. The paper claims state-of-the-art performance across all four datasets.

Significance. If the empirical results are reproducible, DLGAN offers a simple and plausible architectural recipe: extract temporal features from real sequences rather than from random noise, and add a second adversarial layer at the reconstruction stage. This directly targets a known weakness of GAN time-series models and the reported margins in Table 1 are large. The ablation design is reasonable and the full model is best on all four datasets. However, the contribution is entirely empirical; the manuscript provides no code, no data splits, no hyperparameters, no random seeds, and no variance estimates. The main strength is an internally coherent architecture and a plausible training procedure, but the central 'outperforms state-of-the-art' claim cannot currently be verified from the manuscript alone.

major comments (3)

[§5.1, Table 1] The central empirical claim is not verifiable because the evaluation protocol is underspecified. The manuscript reports no preprocessing (normalization, sequence length T, patch/window size), no train/validation/test splits, no hyperparameters, no random seeds, and no confidence intervals. Baseline numbers are reported without variance; for example, the TimeGAN Stock discriminative score (0.102) appears consistent with values in the original TimeGAN paper, while ETTH, Exchange, and Weather were not part of that evaluation. Without a controlled rerun of all baselines under identical splits and preprocessing, or without explicit citations to matching published settings, the margins (e.g., DLGAN 0.079 vs TimeGAN 0.106 on ETTH; 0.173 vs 0.412 on Weather) may reflect protocol differences rather than model quality. The predictive score also requires a defined train/test split (train on synthet
[§4, Eq. (14)] The generator objective includes a supervised reconstruction loss L_mse(H_real, Ĥ_real) on the real hidden sequence. This means Generator2 is directly trained to reconstruct real training data, not only to fool discriminators. The paper should clarify what data are used for this supervised loss and for the downstream discriminative/predictive evaluation. If the same real sequences are used for both training and evaluation, the reported predictive scores could be optimistic because the synthetic data may encode training-set specific information. This is not a circularity in the comparison against external baselines, but the leakage controls must be documented before the superiority claim can be assessed.
[§5.3, Table 2] The ablation results do not consistently support the claim that each component is individually necessary. For instance, on Exchange, w/o All (discriminative 0.127) is better than w/o Temporal Feature Extractor (0.155) and w/o TimeSeries Reconstructor (0.163); on Weather, w/o All (0.177) is nearly as good as the full model (0.173). Similar non-monotonicities appear in the predictive scores. Section 5.3 acknowledges this but provides only a post hoc explanation. The conclusion states that 'ablation studies validate the effectiveness of each module,' which is stronger than the evidence in Table 2 supports. This does not invalidate the main comparison, but it should be presented more cautiously and, ideally, with repeated-seed statistics.

minor comments (4)

[§3.4, Eq. (11)] Equation (11) reads y2 = Discriminator2(H_real_1:T, H_real_1:T, H_fake_1:T); the second argument is likely a typo for Ĥ_real_1:T, based on the text in §3.4. Please correct and specify how the three inputs are combined (e.g., concatenation).
[§5.1, Dataset list] Typo: 'Excahnge' should be 'Exchange'.
[Tables 1 and 2] Tables use 'Methed' instead of 'Method'. Also, the table headers and caption formatting should follow the conference style.
[§3.3, Figure 2 caption] The text refers to 'Figure2(c.' with inconsistent punctuation; make the figure/panel references consistent throughout.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; DLGAN is a straightforward architecture-plus-empirical-evaluation paper with no self-citations and no construction-by-definition.

full rationale

The paper's central claim is empirical: DLGAN outperforms six time-series generation baselines on discriminative and predictive scores across four public datasets (Table 1). Nothing in the derivation chain makes these results true by construction. The model is defined by an encoder/decoder autoencoder (Eqs. 1–2), a temporal feature extractor and generator (Eqs. 3–8), and a reconstructor (Eqs. 9–11), trained with the supervised MSE losses (Eqs. 12–13) and GAN losses (Eqs. 14–15). The supervised reconstruction loss L_mse(H_real, H_real_hat) is an intended training objective that fits the generator to the real hidden sequence; it is not a metric being relabeled as a prediction. The discriminative and predictive scores come from TimeGAN's evaluation protocol applied to external baselines, not from any fitted parameter of DLGAN. There are no self-citations, no imported uniqueness theorems, and no ansatz smuggled in via citation. The main weakness is reproducibility reporting: preprocessing, splits, sequence lengths, seeds, and confidence intervals are absent, so the exact margins over baselines cannot be independently verified from the manuscript. That is a reporting/evidence-quality concern, not circularity. Therefore the appropriate circularity score is 0.

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

The central empirical claim rests on unspecified architecture choices (latent dimension, patch size, all training hyperparameters) and on an evaluation protocol whose comparability across baselines is assumed rather than demonstrated. No new theoretical entities are introduced, which limits the paper's ambition but also means no unjustified physical or conceptual inventions.

free parameters (3)

autoencoder latent dimension N
The number of hidden attributes N in the encoder output is chosen by hand and not reported. This controls the bottleneck through which all temporal features must pass.
patch/window size in temporal feature extractor
The sliding window size in PAT CH is 'appropriately sized' but never specified, and it directly controls what temporal dynamics the feature extractor sees.
all GAN/GRU/attention hyperparameters
Number of GRU layers, hidden sizes, number of attention heads, learning rates, and training iterations are omitted entirely; the reported results depend on these choices.

assumptions (3)

domain assumption The TimeGAN evaluation protocol (discriminative score, predictive score, t-SNE) is a valid measure of synthetic time series quality.
The paper adopts the metrics from the TimeGAN paper without discussing their limitations, such as their insensitivity on near-degenerate predictive tasks.
domain assumption Preprocessing and evaluation of the four public datasets match the baselines' original settings.
Section 5.1 lists datasets by name only. No normalization, sequence length, or split details are given, so the comparison assumes de facto comparability.
domain assumption Standard GAN training with MSE reconstruction loss converges and the two-stage pretraining is stable at the reported hyperparameters.
This is the training procedure described in Section 4; the paper reports no training curves, no convergence analysis, and no seed variance.

how reviews work

0 comments

Cite this review

Pith. "Pith review of DLGAN : Time Series Synthesis Based on Dual-Layer Generative Adversarial Networks." pith.science (2026). https://pith.science/paper/GMEXT3I5

@misc{pith2026250821340,
  author       = {Pith},
  title        = {Pith review of: DLGAN : Time Series Synthesis Based on Dual-Layer Generative Adversarial Networks},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GMEXT3I5}},
  note         = {Machine review of arXiv:2508.21340}
}

read the original abstract

Time series synthesis is an effective approach to ensuring the secure circulation of time series data. Existing time series synthesis methods typically perform temporal modeling based on random sequences to generate target sequences, which often struggle to ensure the temporal dependencies in the generated time series. Additionally, directly modeling temporal features on random sequences makes it challenging to accurately capture the feature information of the original time series. To address the above issues, we propose a simple but effective generative model \textbf{D}ual-\textbf{L}ayer \textbf{G}enerative \textbf{A}dversarial \textbf{N}etworks, named \textbf{DLGAN}. The model decomposes the time series generation process into two stages: sequence feature extraction and sequence reconstruction. First, these two stages form a complete time series autoencoder, enabling supervised learning on the original time series to ensure that the reconstruction process can restore the temporal dependencies of the sequence. Second, a Generative Adversarial Network (GAN) is used to generate synthetic feature vectors that align with the real-time sequence feature vectors, ensuring that the generator can capture the temporal features from real time series. Extensive experiments on four public datasets demonstrate the superiority of this model across various evaluation metrics.

Figures

Figures reproduced from arXiv: 2508.21340 by the authors.

**Figure 1.** Overall architecture of the DLGAN the original time-series data 𝑋1:𝑇 = (𝑋1, · · · , 𝑋𝑇 ), where 𝑋𝑡 = (𝑋 1 𝑡 , 𝑋2 𝑡 , · · · , 𝑋 𝑀 𝑡 ) represents the feature vector of the multidimensional time series at time step 𝑡, and 𝑀 is the number of attributes contained in the time series. The encoder generates a hidden sequence 𝐻 𝑅𝑒𝑎𝑙 1:𝑇 = (𝐻 𝑅𝑒𝑎𝑙 1 , 𝐻𝑅𝑒𝑎𝑙 2 , · · · , 𝐻𝑅𝑒𝑎𝑙 𝑇 ), where 𝐻 𝑅𝑒𝑎𝑙 𝑡 = (𝐻 1 𝑡 , 𝐻2 𝑡 , · · · , 𝐻𝑁 𝑡… view at source ↗

**Figure 2.** Module Detailed Architecture Diagram time series, enabling the capture of correlations between different information channels at the same time step. In channel-independent modeling, since the temporal features of a time series are often reflected in its dynamic changes, the value at a single time step does not carry rich temporal semantics[24, 44]. Therefore, for the input univariate time series, we select an appro… view at source ↗

**Figure 3.** t-SNE visualization. Red denotes original data, and blue denotes synthetic data. [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

45 extracted references · 33 canonical work pages

[1]

Sensegen: A deep learning architecture for synthetic sensor data generation

Moustafa Alzantot, Supriyo Chakraborty, and Mani Srivastava. Sensegen: A deep learning architecture for synthetic sensor data generation. In 2017 IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops), pages 188–193. IEEE, 2017

work page 2017
[2]

Synthsonic: Fast, probabilistic modeling and synthesis of tabular data

Max Baak, Simon Brugman, Ilan Fridman Rojas, Lorraine Dalmeida, Ralph EQ Urlus, and Jean-Baptiste Oger. Synthsonic: Fast, probabilistic modeling and synthesis of tabular data. In International Conference on Artificial Intelligence and Statistics, pages 4747–4763. PMLR, 2022

work page 2022
[3]

Data synthesis via differentially private markov random fields.Proceedings of the VLDB Endowment, 14(11):2190–2202, 2021

Kuntai Cai, Xiaoyu Lei, Jianxin Wei, and Xiaokui Xiao. Data synthesis via differentially private markov random fields.Proceedings of the VLDB Endowment, 14(11):2190–2202, 2021

2021
[4]

Gan-leaks: A taxonomy of membership inference attacks against generative models

Dingfan Chen, Ning Yu, Yang Zhang, and Mario Fritz. Gan-leaks: A taxonomy of membership inference attacks against generative models. In Proceedings of the 2020 ACM SIGSAC conference on computer and communications security , pages 343–362, 2020

work page 2020
[5]

Locally differentially private high-dimensional data synthesis.Science China Information Sciences, 66(1):112101, 2023

Xue Chen, Cheng Wang, Qing Yang, Teng Hu, and Changjun Jiang. Locally differentially private high-dimensional data synthesis.Science China Information Sciences, 66(1):112101, 2023

work page 2023
[6]

Em- pirical evaluation of gated recurrent neural networks on sequence modeling

Junyoung Chung, Caglar Gulcehre, Kyunghyun Cho, and Yoshua Bengio. Em- pirical evaluation of gated recurrent neural networks on sequence modeling. In NIPS 2014 Workshop on Deep Learning, December 2014 , 2014. Conference’17, July 2017, Washington, DC, USA Hou X., Liu S. et al

work page 2014
[7]

Timevae: A variational auto-encoder for multivariate time series generation

Abhyuday Desai, Cynthia Freeman, Zuhui Wang, and Ian Beaver. Timevae: A variational auto-encoder for multivariate time series generation. arXiv preprint arXiv:2111.08095, 2021

arXiv 2021
[8]

Adversarial audio synthe- sis

Chris Donahue, Julian McAuley, and Miller Puckette. Adversarial audio synthe- sis. In International Conference on Learning Representations , 2019

work page 2019

Show all 45 references

[9]

Adarnn: Adaptive learning and forecasting of time series

Yuntao Du, Jindong Wang, Wenjie Feng, Sinno Pan, Tao Qin, Renjun Xu, and Chongjun Wang. Adarnn: Adaptive learning and forecasting of time series. In Proceedings of the 30th ACM international conference on information & knowledge management, pages 402–411, 2021

2021
[10]

Real-valued (med- ical) time series generation with recurrent conditional gans

Cristóbal Esteban, Stephanie L Hyland, and Gunnar Rätsch. Real-valued (med- ical) time series generation with recurrent conditional gans. arXiv preprint arXiv:1706.02633, 2017

2017 arXiv
[11]

Relational data synthesis using generative adversarial networks

Ju Fan, Junyou Chen, Tongyu Liu, Yuwei Shen, Guoliang Li, and Xiaoyong Du. Relational data synthesis using generative adversarial networks. Proceedings of the VLDB Endowment, 13(12):1962–1975, 2020

1962
[12]

Generative adversarial nets

Ian Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio. Generative adversarial nets. Advances in neural information processing systems , 27, 2014

2014
[13]

The capacity and robustness trade- off: Revisiting the channel independent strategy for multivariate time series forecasting

Lu Han, Han-Jia Ye, and De-Chuan Zhan. The capacity and robustness trade- off: Revisiting the channel independent strategy for multivariate time series forecasting. IEEE Transactions on Knowledge and Data Engineering , 2024

2024
[14]

Denoising diffusion probabilistic models

Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising diffusion probabilistic models. Advances in neural information processing systems , 33:6840–6851, 2020

2020
[15]

Long short-term memory

Sepp Hochreiter and Jürgen Schmidhuber. Long short-term memory. Neural computation, 9(8):1735–1780, 1997

1997
[16]

Psa-gan: Progressive self attention gans for synthetic time series

Paul Jeha, Michael Bohlke-Schneider, Pedro Mercado, Shubham Kapoor, Ra- jbir Singh Nirwan, Valentin Flunkert, Jan Gasthaus, and Tim Januschowski. Psa-gan: Progressive self attention gans for synthetic time series. In Interna- tional Conference on Learning Representations , 2022

2022
[17]

Progressive growing of gans for improved quality, stability, and variation

Tero Karras, Timo Aila, Samuli Laine, and Jaakko Lehtinen. Progressive growing of gans for improved quality, stability, and variation. InInternational Conference on Learning Representations, 2018

2018
[18]

Auto-encoding variational bayes

Diederik P Kingma and Max Welling. Auto-encoding variational bayes. arXiv preprint arXiv:1312.6114, 2013

2013 arXiv
[19]

Modeling long-and short-term temporal patterns with deep neural networks

Guokun Lai, Wei-Cheng Chang, Yiming Yang, and Hanxiao Liu. Modeling long-and short-term temporal patterns with deep neural networks. In The 41st international ACM SIGIR conference on research & development in information retrieval, pages 95–104, 2018

2018
[20]

In- vertible tabular gans: Killing two birds with one stone for tabular data synthesis

Jaehoon Lee, Jihyeon Hyeong, Jinsung Jeon, Noseong Park, and Jihoon Cho. In- vertible tabular gans: Killing two birds with one stone for tabular data synthesis. Advances in Neural Information Processing Systems , 34:4263–4273, 2021

2021
[21]

Causal recurrent variational autoencoder for medical time series generation

Hongming Li, Shujian Yu, and Jose Principe. Causal recurrent variational autoencoder for medical time series generation. In Proceedings of the AAAI conference on artificial intelligence , volume 37, pages 8562–8570, 2023

2023
[22]

A critical review of recurrent neural networks for sequence learning.arXiv preprint arXiv:1506.00019, 2015

Zachary C Lipton, John Berkowitz, and Charles Elkan. A critical review of recurrent neural networks for sequence learning.arXiv preprint arXiv:1506.00019, 2015

2015 arXiv
[23]

C-rnn-gan: Continuous recurrent neural networks with adversar- ial training

Olof Mogren. C-rnn-gan: Continuous recurrent neural networks with adversar- ial training. arXiv preprint arXiv:1611.09904, 2016

2016 arXiv
[24]

A time series is worth 64 words: Long-term forecasting with transformers

Yuqi Nie, Nam H Nguyen, Phanwadee Sinthong, and Jayant Kalagnanam. A time series is worth 64 words: Long-term forecasting with transformers. In The Eleventh International Conference on Learning Representations , 2023

2023
[25]

Data synthesis based on generative adversarial networks

Noseong Park, Mahmoud Mohammadi, Kshitij Gorde, Sushil Jajodia, Hongkyu Park, and Youngmin Kim. Data synthesis based on generative adversarial networks. Proceedings of the VLDB Endowment , 11(10), 2018

2018
[26]

Unsupervised representa- tion learning with deep convolutional generative adversarial networks

Alec Radford, Luke Metz, and Soumith Chintala. Unsupervised representa- tion learning with deep convolutional generative adversarial networks. arXiv preprint arXiv:1511.06434, 2015

2015 arXiv
[27]

Dae-gan: Dynamic aspect-aware gan for text-to-image synthesis

Shulan Ruan, Yong Zhang, Kun Zhang, Yanbo Fan, Fan Tang, Qi Liu, and Enhong Chen. Dae-gan: Dynamic aspect-aware gan for text-to-image synthesis. In Proceedings of the IEEE/CVF International Conference on Computer Vision , pages 13960–13969, 2021

2021
[28]

Scale- former: Iterative multi-scale refining transformers for time series forecasting

Mohammad Amin Shabani, Amir H Abdi, Lili Meng, and Tristan Sylvain. Scale- former: Iterative multi-scale refining transformers for time series forecasting. In International Conference on Learning Representations , 2023

2023
[29]

Multi-resolution diffusion models for time series forecasting

Lifeng Shen, Weiyu Chen, and James Kwok. Multi-resolution diffusion models for time series forecasting. In The Twelfth International Conference on Learning Representations, 2024

2024
[30]

P3gm: Private high-dimensional data release via privacy preserving phased generative model

Shun Takagi, Tsubasa Takahashi, Yang Cao, and Masatoshi Yoshikawa. P3gm: Private high-dimensional data release via privacy preserving phased generative model. In 2021 IEEE 37th International Conference on Data Engineering (ICDE) , pages 169–180. IEEE, 2021

2021
[31]

WaveNet: A Generative Model for Raw Audio

Aäron van den Oord, Sander Dieleman, Heiga Zen, Karen Simonyan, Oriol Vinyals, Alex Graves, Nal Kalchbrenner, Andrew Senior, and Koray Kavukcuoglu. WaveNet: A Generative Model for Raw Audio. In Proc. 9th ISCA Workshop on Speech Synthesis Workshop (SSW 9) , page 125, 2016

2016
[32]

Visualizing data using t-sne

Laurens Van der Maaten and Geoffrey Hinton. Visualizing data using t-sne. Journal of machine learning research , 9(11), 2008

2008
[33]

Diffusion-gan: Training gans with diffusion

Zhendong Wang, Huangjie Zheng, Pengcheng He, Weizhu Chen, and Mingyuan Zhou. Diffusion-gan: Training gans with diffusion. In International Conference on Learning Representations, 2023

2023
[34]

Transformers in time series: a survey

Qingsong Wen, Tian Zhou, Chaoli Zhang, Weiqi Chen, Ziqing Ma, Junchi Yan, and Liang Sun. Transformers in time series: a survey. InProceedings of the Thirty- Second International Joint Conference on Artificial Intelligence , pages 6778–6786, 2023

2023
[35]

Autoformer: Decom- position transformers with auto-correlation for long-term series forecasting

Haixu Wu, Jiehui Xu, Jianmin Wang, and Mingsheng Long. Autoformer: Decom- position transformers with auto-correlation for long-term series forecasting. Advances in neural information processing systems , 34:22419–22430, 2021

2021
[36]

Lightweight privacy-preserving gan framework for model training and image synthesis

Yang Yang, Ke Mu, and Robert H Deng. Lightweight privacy-preserving gan framework for model training and image synthesis. IEEE Transactions on Infor- mation Forensics and Security , 17:1083–1098, 2022

2022
[37]

Time-series generative adversarial networks

Jinsung Yoon, Daniel Jarrett, and Mihaela Van der Schaar. Time-series generative adversarial networks. Advances in neural information processing systems , 32, 2019

2019
[38]

Learning to learn the future: Modeling concept drifts in time series prediction

Xiaoyu You, Mi Zhang, Daizong Ding, Fuli Feng, and Yuanmin Huang. Learning to learn the future: Modeling concept drifts in time series prediction. In Pro- ceedings of the 30th ACM International Conference on Information & Knowledge Management, pages 2434–2443, 2021

2021
[39]

Ds- former: A double sampling transformer for multivariate time series long-term prediction

Chengqing Yu, Fei Wang, Zezhi Shao, Tao Sun, Lin Wu, and Yongjun Xu. Ds- former: A double sampling transformer for multivariate time series long-term prediction. In Proceedings of the 32nd ACM International Conference on Informa- tion and Knowledge Management , pages 3062–3072, 2023

2023
[40]

Attributing fake images to gans: Learn- ing and analyzing gan fingerprints

Ning Yu, Larry S Davis, and Mario Fritz. Attributing fake images to gans: Learn- ing and analyzing gan fingerprints. In Proceedings of the IEEE/CVF international conference on computer vision , pages 7556–7566, 2019

2019
[41]

Diffusion-TS: Interpretable diffusion for general time series generation

Xinyu Yuan and Yan Qiao. Diffusion-TS: Interpretable diffusion for general time series generation. In The Twelfth International Conference on Learning Representations, 2024

2024
[42]

Quality-aware self-training on differentiable synthesis of rare relational data

Chongsheng Zhang, Yaxin Hou, Ke Chen, Shuang Cao, Gaojuan Fan, and Ji Liu. Quality-aware self-training on differentiable synthesis of rare relational data. In Proceedings of the AAAI Conference on Artificial Intelligence , volume 37, pages 6602–6611, 2023

2023
[43]

Privbayes: Private data release via bayesian networks

Jun Zhang, Graham Cormode, Cecilia M Procopiuc, Divesh Srivastava, and Xiaokui Xiao. Privbayes: Private data release via bayesian networks. ACM Transactions on Database Systems (TODS) , 42(4):1–41, 2017

2017
[44]

Crossformer: Transformer utilizing cross- dimension dependency for multivariate time series forecasting

Yunhao Zhang and Junchi Yan. Crossformer: Transformer utilizing cross- dimension dependency for multivariate time series forecasting. In The eleventh international conference on learning representations , 2022

2022
[45]

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

Haoyi Zhou, Shanghang Zhang, Jieqi Peng, Shuai Zhang, Jianxin Li, Hui Xiong, and Wancai Zhang. Informer: Beyond efficient transformer for long sequence time-series forecasting. In Proceedings of the AAAI conference on artificial intelligence, volume 35, pages 11106–11115, 2021

2021

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.

[1] [1]

Sensegen: A deep learning architecture for synthetic sensor data generation

Moustafa Alzantot, Supriyo Chakraborty, and Mani Srivastava. Sensegen: A deep learning architecture for synthetic sensor data generation. In 2017 IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops), pages 188–193. IEEE, 2017

work page 2017

[2] [2]

Synthsonic: Fast, probabilistic modeling and synthesis of tabular data

Max Baak, Simon Brugman, Ilan Fridman Rojas, Lorraine Dalmeida, Ralph EQ Urlus, and Jean-Baptiste Oger. Synthsonic: Fast, probabilistic modeling and synthesis of tabular data. In International Conference on Artificial Intelligence and Statistics, pages 4747–4763. PMLR, 2022

work page 2022

[3] [3]

Data synthesis via differentially private markov random fields.Proceedings of the VLDB Endowment, 14(11):2190–2202, 2021

Kuntai Cai, Xiaoyu Lei, Jianxin Wei, and Xiaokui Xiao. Data synthesis via differentially private markov random fields.Proceedings of the VLDB Endowment, 14(11):2190–2202, 2021

2021

[4] [4]

Gan-leaks: A taxonomy of membership inference attacks against generative models

Dingfan Chen, Ning Yu, Yang Zhang, and Mario Fritz. Gan-leaks: A taxonomy of membership inference attacks against generative models. In Proceedings of the 2020 ACM SIGSAC conference on computer and communications security , pages 343–362, 2020

work page 2020

[5] [5]

Locally differentially private high-dimensional data synthesis.Science China Information Sciences, 66(1):112101, 2023

Xue Chen, Cheng Wang, Qing Yang, Teng Hu, and Changjun Jiang. Locally differentially private high-dimensional data synthesis.Science China Information Sciences, 66(1):112101, 2023

work page 2023

[6] [6]

Em- pirical evaluation of gated recurrent neural networks on sequence modeling

Junyoung Chung, Caglar Gulcehre, Kyunghyun Cho, and Yoshua Bengio. Em- pirical evaluation of gated recurrent neural networks on sequence modeling. In NIPS 2014 Workshop on Deep Learning, December 2014 , 2014. Conference’17, July 2017, Washington, DC, USA Hou X., Liu S. et al

work page 2014

[7] [7]

Timevae: A variational auto-encoder for multivariate time series generation

Abhyuday Desai, Cynthia Freeman, Zuhui Wang, and Ian Beaver. Timevae: A variational auto-encoder for multivariate time series generation. arXiv preprint arXiv:2111.08095, 2021

arXiv 2021

[8] [8]

Adversarial audio synthe- sis

Chris Donahue, Julian McAuley, and Miller Puckette. Adversarial audio synthe- sis. In International Conference on Learning Representations , 2019

work page 2019