Pith. sign in

REVIEW 4 major objections 4 minor 42 references

A Comparative Study of Open-Source Libraries for Synthetic Tabular Data Generation: SDV vs. SynthCity

T0 review · 4 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read In a six-way comparison on 1,000 energy-data rows, Synthcity's Bayesian Network reproduced the real tables most faithfully, while SDV's TVAE led when asked to generate 10,000 rows.

desk verdict Useful head-to-head benchmark, but the missing target variable and sampling procedure make the headline ranking unverifiable as written. read the letter →

arxiv 2506.17847 v1 pith:3ZVQSJPD submitted 2025-06-21 cs.LG cs.AI

Cristian Del Gobbo This is my paper · ORCID
classification cs.LGcs.AI
keywords syntheticdatagenerationtabularSDVSynthcityBayesiannetworksCTGANTVAEpredictiveutility
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 tries to establish, through a controlled comparison on a real energy-consumption dataset, that when the goal is statistical fidelity of synthetic tabular data, Synthcity's Bayesian Network is the strongest of six generators from two open-source libraries, and that when the goal is predictive utility at tenfold expansion, SDV's TVAE is strongest. The study matters because small organizations with limited real data need guidance on which generator to trust. The paper also argues that no library has an overall performance edge, so usability and documentation become the deciding factor.

What carries the argument

The machinery is the evaluation protocol rather than a single generative model. Statistical similarity is a column-wise composite score that averages normalized differences in mean, median, and standard deviation, plus the Kolmogorov–Smirnov statistic and Wasserstein distance for numeric columns and mode agreement for categorical columns. Predictive utility uses the Train-on-Synthetic, Test-on-Real (TSTR) paradigm: four regression models (XGBRegressor, Random Forest Regressor, Support Vector Regressor, and Linear Regression) are trained on each synthetic dataset and tested on real data, then scored relative to the same models trained on real data. That scoring protocol is what carries all the paper's rankings.

What would settle it

Re-run the TSTR evaluation on the same energy dataset with the regression target explicitly named and with the 1,000 training rows drawn by stratified random sampling. If SDV's TVAE no longer produces the best 1:10 predictive utility score, or if Synthcity's Bayesian Network keeps that lead, the paper's headline ranking reverses.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is a ranking: Synthcity's Bayesian Network produces tables most statistically like the real data in both the 1:1 setting (fidelity score 96.53 out of 100) and the 1:10 setting (78.25), while SDV's TVAE produces synthetic rows with the highest predictive utility when expanding from 1,000 to 10,000 rows (score 0.31), where most other generators score near or below zero. The paper also finds that statistical similarity stays acceptably high across all generators as output scales, but predictive utility falls sharply, and that the two libraries' shared models behave differently, with SDV's CTGAN and TVAE outperforming their Synthcity counterparts in the 1:10 predictive test.

Load-bearing premise

The predictive-utility ranking rests on an unstated premise: that the regression target column among the dataset's 29 features and the way the 1,000 training rows were sampled are exactly what the evaluator assumed, but the paper never names the target or the sampling procedure.

Editorial extensions

If this is right

  • A practitioner who needs statistically plausible tables from a small sample should reach for Synthcity's Bayesian Network before trying deep generators.
  • A practitioner who needs to produce ten times more rows than they have and wants those rows to preserve predictive value should reach for SDV's TVAE.
  • Generating at a 1:10 scale appears safe for distributional shape (fidelity scores 70–80) but risky for downstream modeling (most predictive-utility scores are negative), so the scale ratio is a decision variable, not a default.
  • Because the paper finds no overall performance winner, library choice can reasonably be driven by documentation and community support, where SDV is judged stronger.

Reading between the lines

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

  • Editorial inference: the Bayesian Network's dominance likely reflects the dataset's strong inter-column dependencies (energy consumption with weather and house sensors), so the ranking may not transfer to datasets with weak or nonlinear dependencies; a useful next test is repeating the protocol on a high-cardinality categorical dataset.
  • Editorial inference: the TSTR scoring's sensitivity to the unstated regression target means the predictive-utility ranking is less robust than the fidelity ranking; re-running with the target column declared and stratified sampling of the 1,000 training rows is a cheap falsification.
  • Editorial inference: the sharp drop in predictive utility at 1:10 while fidelity stays moderate suggests the fidelity metric rewards matching marginal distributions but fails to capture joint dependencies that matter for prediction; this motivates adding a dependence-preservation metric.
  • Editorial inference: a practical extension would be to test whether training the generators on 2,000 or 5,000 rows changes the 1:10 ranking, since the paper's 1,000-row sample may understate deep models' behavior with more data.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 4 minor

Summary. This manuscript compares six tabular synthetic data generators drawn from two open-source libraries: Gaussian Copula, CTGAN, and TVAE from SDV, and Bayesian Network, CTGAN, and TVAE from Synthicity. The experiments use a UCI energy-consumption dataset, train each generator on 1,000 real rows, and generate synthetic data at 1:1 (1,000 rows) and 1:10 (10,000 rows) ratios. The generators are evaluated on statistical similarity and on predictive utility via a Train on Synthetic, Test on Real (TSTR) protocol with four regression models. The reported results are that Synthicity's Bayesian Network achieves the highest statistical fidelity in both scenarios, that SDV's TVAE performs best for predictive utility in the 1:10 setting, and that neither library is an overall performance winner, with SDV preferred on documentation and usability grounds.

Significance. If the claims held as stated, the paper would provide useful practical guidance for practitioners choosing between SDV and Synthicity. The tabulated results show internal consistency: the predictive-utility scores in Tables 5 and 6 are plausibly derived from the normalized MAE, MSE, and R2 differences in Tables 3 and 4, and the TSTR design is methodologically appropriate. However, the manuscript omits several load-bearing experimental details: the regression target column is not identified, the sampling of the 1,000-row training subset is not described, the weights of the custom statistical-similarity metric are not given, and no repository URL is provided. As written, the headline rankings are not reproducible from the paper, so the practical guidance cannot currently be verified.

major comments (4)
  1. [Methods: Comparison Process and Methods] The predictive-utility evaluation never states which of the 29 columns of the UCI energy dataset is the regression target, nor how the 1,000-row training subset was sampled (random, stratified, chronological, or fixed by a seed). Tables 5 and 6 are normalized differences relative to models trained on real data, so both choices affect every entry; the claim that TVAE (SDV) is the best 1:10 model is therefore not verifiable from the text as written.
  2. [Methods: Comparison Process and Methods] The statistical-similarity score is defined only as 'a weighted combination' of normalized mean/median/std, Kolmogorov-Smirnov, and Wasserstein differences, with no weights or aggregation details given. Tables 1 and 2, which support the headline claim that Synthicity's Bayesian Network has the highest fidelity in both settings, are rankings of this unspecified score; the weights are free parameters that should be reported.
  3. [Methods: Comparison Process and Methods] The text states that 'multiple evaluation rounds were conducted using repeated holdout validation' but gives no number of rounds, train/test split ratio, seeds, or number of synthetic-generation runs; it also gives no hyperparameters or package versions, and the 'accompanying GitHub repository' is not identified by URL. The SDV-versus-Synthcity differences in Tables 3 and 4 could be within stochastic variation, so the results cannot be distinguished from noise or independently checked as reported.
  4. [Results: Table 4 note] The note says that 'different amounts of real data were used for training' in the 1:1 and 1:10 experiments, but the paper does not state whether the 10,000 real training rows in the 1:10 condition are a superset of the 1,000 rows used to train the generators or an independent draw. This is essential for interpreting the comparison between models trained on 1:10 synthetic data and models trained on real data.
minor comments (4)
  1. [Methods: Data Generators Description] The section says 'four different Data Generators' although six models are evaluated; please clarify that four architectures are instantiated by two libraries.
  2. [Literature Review] The literature review describes statistical similarity as using 'reliable statistical functions provided by SDV', while the Methods and Abstract describe a 'custom' function; specify which components come from SDV and which are custom, since using a library under comparison inside the evaluation metric is a potential fairness concern.
  3. [Discussion & Conclusion] The phrase 'no significant distinction' and 'no significant performance gap' are used without reporting any significance test; please either provide a test or replace 'significant' with a non-statistical term such as 'substantive'.
  4. [General] The title uses 'SynthCity' while the body uses 'Synthicity' inconsistently; standardize the spelling for consistency.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the benchmark conclusions are evaluated against an external real-world dataset and do not reduce to their inputs by construction.

full rationale

This paper does not attempt a theoretical derivation; it reports an empirical benchmark of six synthetic-data generators against the UCI energy dataset, an external real-world source. The statistical-similarity and predictive-utility scores are computed from held-out real data via standard metrics (KS statistic, Wasserstein distance, MAE/MSE/R2 in a TSTR setup), so the central claims are falsifiable by re-running the comparison rather than being forced by definition. The only mild self-reference is that some statistical functions are 'provided by SDV,' one of the libraries under comparison, but the paper describes the scoring as a custom column-by-column computation of standard statistics, and using a library's implementation of generic statistical tests is a fairness or implementation concern, not a circular reduction of the conclusion to its input. Similarly, the unstated regression target and 1,000-row sampling procedure are reproducibility risks that make the predictive-utility ranking hard to verify, but they do not constitute a self-definitional or fitted-input-called-prediction step. There are no uniqueness theorems imported from the authors, no ansatz smuggled in via self-citation, and no renaming of a known result; the conclusions are presented as measurements, not as implications that reproduce their own premises. Accordingly, no circularity step meets the evidentiary bar, and the appropriate score is 0.

Assumptions & free parameters 1 free parameters · 6 assumptions · 0 invented entities

The paper introduces no new theoretical entities or fitted constants. Its conclusions rest on the experimental assumptions listed above, most importantly the unstated target variable, subset sampling, and default hyperparameters. The similarity-score weights are a hand-chosen free parameter reported nowhere.

free parameters (1)
  • Column-score weights in statistical similarity function = not reported
    The custom metric combines mean, median, standard deviation, KS statistic, Wasserstein distance, and categorical mode with unspecified weights. The 0-100 similarity scores in Tables 1 and 2 depend on these unpublished weights.
assumptions (6)
  • standard math Sklar's theorem (Eq. 1) and Bayesian network factorization (Eq. 2) are valid for the data.
    Used to describe Gaussian Copula and BN generators; standard results from cited literature, not at issue.
  • standard math The VAE ELBO objective (Eq. 3) describes TVAE training.
    Background description of TVAE; taken from Xu et al. [31].
  • domain assumption The 1,000-row subset represents the full 19,735-row energy dataset.
    The paper says a subset is used but does not state the sampling method; if the subset is biased, all fidelity and utility rankings could change.
  • domain assumption The regression target variable is the one the evaluator intended.
    The TSTR section never names the target column; the predictive utility tables depend on this unstated choice.
  • domain assumption Each generator's default hyperparameters are representative of typical library performance.
    No tuning or sensitivity analysis is reported; the paper itself ascribes performance differences to Synthcity's deeper networks and different training schemes, so results may be configuration-specific.
  • domain assumption The custom statistical similarity score captures fidelity.
    Categorical columns are scored only by mode matching and numerical columns by an undisclosed weighted combination of statistics, so the 0-100 scores may not reflect true distributional similarity.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Comparative Study of Open-Source Libraries for Synthetic Tabular Data Generation: SDV vs. SynthCity." pith.science (2026). https://pith.science/paper/3ZVQSJPD

@misc{pith2026250617847,
  author       = {Pith},
  title        = {Pith review of: A Comparative Study of Open-Source Libraries for Synthetic Tabular Data Generation: SDV vs. SynthCity},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3ZVQSJPD}},
  note         = {Machine review of arXiv:2506.17847}
}
read the original abstract

High-quality training data is critical to the performance of machine learning models, particularly Large Language Models (LLMs). However, obtaining real, high-quality data can be challenging, especially for smaller organizations and early-stage startups. Synthetic data generators provide a promising solution by replicating the statistical and structural properties of real data while preserving privacy and scalability. This study evaluates the performance of six tabular synthetic data generators from two widely used open-source libraries: SDV (Gaussian Copula, CTGAN, TVAE) and Synthicity (Bayesian Network, CTGAN, TVAE). Using a real-world dataset from the UCI Machine Learning Repository, comprising energy consumption and environmental variables from Belgium, we simulate a low-data regime by training models on only 1,000 rows. Each generator is then tasked with producing synthetic datasets under two conditions: a 1:1 (1,000 rows) and a 1:10 (10,000 rows) input-output ratio. Evaluation is conducted using two criteria: statistical similarity, measured via classical statistics and distributional metrics; and predictive utility, assessed using a "Train on Synthetic, Test on Real" approach with four regression models. While statistical similarity remained consistent across models in both scenarios, predictive utility declined notably in the 1:10 case. The Bayesian Network from Synthicity achieved the highest fidelity in both scenarios, while TVAE from SDV performed best in predictive tasks under the 1:10 setting. Although no significant performance gap was found between the two libraries, SDV stands out for its superior documentation and ease of use, making it more accessible for practitioners.

Figures

Figures reproduced from arXiv: 2506.17847 by the authors.

Figure 1
Figure 1. TSTR paradigm explained visually. XGBRegressor, Random Forest Regressor, Support Vector Regressor (SVR), and Linear Regression. Critically, the trained models were tested only on subsets of the real data, ensuring a valid and realistic evaluation scenario. Multiple evaluation rounds were conducted using repeated holdout validation, each time varying the training-test split to obtain robust and statistically signific… view at source ↗
Figure 2
Figure 2. Overview of the 1:1 experimental setting for synthetic data generation: [PITH_FULL_IMAGE:figures/full_fig_p013_2.png] view at source ↗
Figure 3
Figure 3. Overview of the 1:10 experimental setting for synthetic data generation: [PITH_FULL_IMAGE:figures/full_fig_p013_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

42 extracted references · 24 canonical work pages

  1. [1]

    Quality of Data in Machine Learning

    Antti Kariluoto, Arto Pärnänen, Joni Kultanen, Jukka Soininen, and Pekka Abrahamsson. Quality of data in machine learning.arXiv preprint arXiv:2112.09400, 2021

    work page Pith review arXiv 2021

  2. [2]

    Beam, Irene Y

    Marzyeh Ghassemi, Tristan Naumann, Peter Schulam, Andrew L. Beam, Irene Y. Chen, Elias Chaibub Neto, and Mihaela van der Schaar. Opportu- nities in machine learning for healthcare.arXiv preprint arXiv:1806.00388, 2018

    arXiv 2018

  3. [3]

    Mihaela Cătălina Stoian, Alexandru Gîrbea, and Anca D. Jurcut. A survey on tabular data generation: Utility, alignment, fidelity, privacy, and beyond. arXiv preprint arXiv:2503.05954, 2025

    arXiv 2025

  4. [4]

    Nikolenko

    Sergey I. Nikolenko. Synthetic data for deep learning. arXiv preprint arXiv:1909.11512, 2019

    arXiv 1909

  5. [5]

    Data Augmentation Approaches in Natural Language Processing: A Survey

    Bohan Li, Hao Zhou, Junxian He, Mingxuan Wang, Yiming Yang, and Lei Li. Data augmentation approaches in natural language processing: A survey. arXiv preprint arXiv:2110.01852, 2021

    work page Pith review arXiv 2021

  6. [6]

    OpenAI. ChatGPT. https://openai.com/index/chatgpt/, 2022. Retrieved from https://openai.com/index/chatgpt/

    work page 2022

  7. [7]

    Shaji George and A

    A. Shaji George and A. S. Hovan George. A review of chatgpt ai’s impact on several business sectors.Partners Universal International Innovation Journal, 1(1):9–23, 2023

    work page 2023

  8. [8]

    Hao Chen, Abdul Waheed, Xiang Li, Yidong Wang, Jindong Wang, Bhiksha Raj, and Marah I. Abdin. On the diversity of synthetic data and its 19 impact on training large language models.arXiv preprint arXiv:2410.15226, October 2024

    arXiv

Show all 42 references
  1. [9]

    Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared D

    Tom B. Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared D. Kaplan, Prafulla Dhariwal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, Sandhini Agarwal, Ariel Herbert-Voss, Gretchen Krueger, Tom Henighan, Rewon Child, Aditya Ramesh, Daniel M. Ziegler, Je...

    1901

  2. [10]

    Gpt-4 technical report

    OpenAI. Gpt-4 technical report. Technical report, OpenAI, March 2023. Accessed: 2025-06-21

    2023

  3. [11]

    Constitutional ai: Harmlessness from ai feedback.arXiv preprint arXiv:2212.08073, 2022

    Yuntao Bai, Saurav Kadavath, Sandipan Kundu, Amanda Askell, Jackson Kernion, Tom Henighan, Tristan Hume, David Krueger, Ben Mann, Jacob DasSarma, et al. Constitutional ai: Harmlessness from ai feedback.arXiv preprint arXiv:2212.08073, 2022

    2022 arXiv

  4. [12]

    Sparks of artificial general intelligence: Early experiments with gpt-4.arXiv preprint arXiv:2303.12712, 2023

    Sébastien Bubeck, Varun Chandrasekaran, Ronen Eldan, Johannes Gehrke, Eric Horvitz, Ece Kamar, Peter Lee, Yuanzhi Lee, Yin Tat Li, Scott Lundberg, Harsha Nori, et al. Sparks of artificial general intelligence: Early experiments with gpt-4.arXiv preprint arXiv:2303.12712, 2023

    2023 arXiv

  5. [13]

    Will we run out of data? an analysis of the limits of scaling datasets in machine learning.arXiv preprint arXiv:2211.04325, 2022

    Pablo Villalobos, Jaime Sevilla, Lennart Heim, Tamay Besiroglu, Marius Hobbhahn, and Anson Ho. Will we run out of data? an analysis of the limits of scaling datasets in machine learning.arXiv preprint arXiv:2211.04325, 2022

    2022 arXiv

  6. [14]

    A review of the challenges with massive web-mined corpora used in large language models pre-training

    Michał Perełkiewicz and Rafał Poświata. A review of the challenges with massive web-mined corpora used in large language models pre-training. arXiv preprint arXiv:2407.07630, 2024

    2024 arXiv

  7. [15]

    Identify, align, and integrate: Match- ing knowledge graphs to commonsense reasoning tasks

    Lisa Bauer and Mohit Bansal. Identify, align, and integrate: Match- ing knowledge graphs to commonsense reasoning tasks. arXiv preprint arXiv:2104.10193, 2021

    2021 arXiv

  8. [16]

    Best practices and lessons learned on synthetic data for language models.arXiv preprint arXiv:2404.07503, 2024

    Ruibo Liu, Dian Yu, Zihan Yuan, Jinchao Zhang, Fei Gao, Rui Shu, Graham Neubig, and Taylor Berg-Kirkpatrick. Best practices and lessons learned on synthetic data for language models.arXiv preprint arXiv:2404.07503, 2024

    2024 arXiv

  9. [17]

    Machine learning for synthetic data generation: A review.arXiv preprint arXiv:2302.04062, 2024

    Yingzhou Lu, Xin Meng, Yu Yao, Cong Liu, Xuefeng Li, Haifeng Yang, Zheng Wang, Bin Hu, and Qingpeng Zhang. Machine learning for synthetic data generation: A review.arXiv preprint arXiv:2302.04062, 2024. 20

    2024 arXiv

  10. [18]

    Comprehensive exploration of synthetic data generation: A survey

    André Bauer, Julian Schäfer, Michael Koch, Hanlin Yang, Samuel Kounev, Clemens Klabunde, Marco Rauh, Niko Sunderhauf, Patrick Esser, Ian Foster, and Dong Kim. Comprehensive exploration of synthetic data generation: A survey. arXiv preprint arXiv:2401.02524, 2024

    2024 arXiv

  11. [19]

    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. InAdvances in Neural Information Processing Systems (NeurIPS), volume 27, pages 2672–2680, 2014

    2014

  12. [20]

    Unsupervised represen- tation learning with deep convolutional generative adversarial networks

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

    2015 arXiv

  13. [21]

    Conditional generative adversarial nets

    Mehdi Mirza and Simon Osindero. Conditional generative adversarial nets. arXiv preprint arXiv:1411.1784, 2014

    2014 arXiv

  14. [22]

    Wasserstein gan

    Martin Arjovsky, Soumith Chintala, and Léon Bottou. Wasserstein gan. arXiv preprint arXiv:1701.07875, 2017

    2017 arXiv

  15. [23]

    Kingma and Max Welling

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

    2013 arXiv

  16. [24]

    Denoising diffusion probabilistic models

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising diffusion probabilistic models. In Advances in Neural Information Processing Systems (NeurIPS), volume 33, pages 6840–6851, 2020

    2020

  17. [25]

    Gomez, Łukasz Kaiser, and Illia Polosukhin

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Łukasz Kaiser, and Illia Polosukhin. Attention is all you need. In Advances in Neural Information Processing Systems (NeurIPS), volume 30, pages 5998–6008, 2017

    2017

  18. [26]

    Comparison of synthetic data generation techniques for control group survival data in oncology clinical trials: Simulation study

    Ippei Akiya, Takuma Ishihara, and Keiichi Yamamoto. Comparison of synthetic data generation techniques for control group survival data in oncology clinical trials: Simulation study. JMIR Medical Informatics, 12:e55118, 2024

    2024

  19. [27]

    Copula Theory and Its Applications, volume 198

    Piotr Jaworski, Fabrizio Durante, Wolfgang Karl Härdle, and Tomasz Rychlik. Copula Theory and Its Applications, volume 198. Springer, New York, 2010

    2010

  20. [28]

    The synthetic data vault

    Neha Patki, Roy Wedge, and Kalyan Veeramachaneni. The synthetic data vault. In2016 IEEE International Conference on Data Science and Advanced Analytics (DSAA), pages 399–410. IEEE, 2016

    2016

  21. [29]

    Jim Young, Sudhakar Iyengar, Mouna Mikou, and Jerome P. Reiter. Using bayesian networks to create synthetic data.Journal of Official Statistics, 25(4):549–567, 2009. 21

    2009

  22. [30]

    Procopiuc, Divesh Srivastava, and Xiaokui Xiao

    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

  23. [31]

    Modeling tabular data using conditional gan

    Lei Xu, Maria Skoularidou, Alfredo Cuesta-Infante, and Kalyan Veera- machaneni. Modeling tabular data using conditional gan. InAdvances in Neural Information Processing Systems (NeurIPS), volume 32, Vancouver, Canada, 2019

    2019

  24. [32]

    Ctab-gan+: Enhancing tabular data synthesis.Frontiers in Big Data, 6, 2024

    Zilong Zhao, Aditya Kunar, and Robert Birke. Ctab-gan+: Enhancing tabular data synthesis.Frontiers in Big Data, 6, 2024

    2024

  25. [33]

    Courville

    Ishaan Gulrajani, Faruk Ahmed, Martin Arjovsky, Vincent Dumoulin, and Aaron C. Courville. Improved training of wasserstein gans. InAdvances in Neural Information Processing Systems, volume 30, pages 5767–5777. Curran Associates, Inc., 2017

    2017

  26. [34]

    Pacgan: The power of two samples in generative adversarial networks

    Zinan Lin, Ashish Khetan, Giulia Fanti, and Sewoong Oh. Pacgan: The power of two samples in generative adversarial networks. InAdvances in Neural Information Processing Systems, volume 31, pages 14953–14963. Curran Associates, Inc., 2018

    2018

  27. [35]

    Rigorous experi- mental analysis of tabular data generated using tvae and ctgan.International Journal of Advanced Computer Science and Applications (IJACSA), 15(4), 2024

    Parul Yadav, Richa Vohra, Parul Jain, and Rajan Gupta. Rigorous experi- mental analysis of tabular data generated using tvae and ctgan.International Journal of Advanced Computer Science and Applications (IJACSA), 15(4), 2024

    2024

  28. [36]

    Candanedo, Véronique Feldheim, and Dominique Deramaix

    Luis M. Candanedo, Véronique Feldheim, and Dominique Deramaix. Data driven prediction models of energy use of appliances in a low-energy house. Energy and Buildings, 140:81–97, 2017

    2017

  29. [37]

    Synthetic data, real errors: How (not) to publish and use synthetic data

    Boris van Breugel and Mihaela van der Schaar. Synthetic data, real errors: How (not) to publish and use synthetic data. arXiv preprint arXiv:2305.09235, 2023

    2023 arXiv

  30. [38]

    Deepenergy: Prediction of appliances energy with long-short term memory recurrent neural network

    Erdenebileg Batbaatar, Hyun Woo Park, Dingkun Li, Keun Ho Ryu, and Jaeho Kim. Deepenergy: Prediction of appliances energy with long-short term memory recurrent neural network. InIntelligent Information and Database Systems, volume 10752 ofLecture Notes in Computer Science, pag...

    2018

  31. [39]

    Bayesian network for predicting energy consumption in schools in flori- anópolis – brazil

    Matheus Soares Geraldi, Mateus Vinicius Bavaresco, and Enedir Ghisi. Bayesian network for predicting energy consumption in schools in flori- anópolis – brazil. InProceedings of Building Simulation 2019: 16th Confer- ence of IBPSA, pages 4188–4195, Rome, Italy, 2019. Internatio...

    2019

  32. [40]

    Synthcity: ctgan_plugin.py and tvae_plugin.py

    SynthCity Contributors. Synthcity: ctgan_plugin.py and tvae_plugin.py. https://github.com/vanderschaarlab/synthcity, 2023. Accessed: 2025-06- 21. 22

    2023

  33. [41]

    CTGAN and TVAE Documentation

    SDV Developers. CTGAN and TVAE Documentation. https://docs.sdv.dev,

  34. [2024]

    Accessed: 2025-06-21. 23

    2025

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.