Pith. sign in

REVIEW 25 cited by

TimeVAE: A Variational Auto-Encoder for Multivariate Time Series Generation

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2111.08095 v3 pith:JNOIRDVY submitted 2021-11-15 cs.LG

Abhyuday Desai , Cynthia Freeman , Zuhui Wang , Ian Beaver This is my paper · ORCID
classification cs.LG
keywords datagenerationarchitecturegeneratedproposedtime-seriesdomaingenerative
verification ladder T0 review T1 audit T2 compute T3 formal

Signed reviews

No signed human review yet.

0 comments
read the original abstract

Recent work in synthetic data generation in the time-series domain has focused on the use of Generative Adversarial Networks. We propose a novel architecture for synthetically generating time-series data with the use of Variational Auto-Encoders (VAEs). The proposed architecture has several distinct properties: interpretability, ability to encode domain knowledge, and reduced training times. We evaluate data generation quality by similarity and predictability against four multivariate datasets. We experiment with varying sizes of training data to measure the impact of data availability on generation quality for our VAE method as well as several state-of-the-art data generation methods. Our results on similarity tests show that the VAE approach is able to accurately represent the temporal attributes of the original data. On next-step prediction tasks using generated data, the proposed VAE architecture consistently meets or exceeds performance of state-of-the-art data generation methods. While noise reduction may cause the generated data to deviate from original data, we demonstrate the resulting de-noised data can significantly improve performance for next-step prediction using generated data. Finally, the proposed architecture can incorporate domain-specific time-patterns such as polynomial trends and seasonalities to provide interpretable outputs. Such interpretability can be highly advantageous in applications requiring transparency of model outputs or where users desire to inject prior knowledge of time-series patterns into the generative model.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 25 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Breaking the Curse with BAND: Nonparametric Distribution Estimation in High Dimensions

    stat.ML 2026-07 conditional novelty 6.0 of 10

    Sparse Bayesian-network factorization plus sparsity-aware regression yields polynomial TV rates for high-dimensional mixed-type distribution estimation, beating classical histogram rates under sparsity.

  2. MIDiff: Tackling Sparsity and Imbalance in Mobile Usage Generation via Multivariate-Imaging Diffusion

    cs.LG 2026-07 conditional novelty 6.0 of 10

    A diffusion model trained on C-GASF images of mobile usage generates synthetic user traces that match real trace statistics far better than prior time-series generative baselines.

  3. RMISC: A Large-scale Real-world Multivariate Corpus for Time Series Foundation Models

    cs.AI 2026-07 conditional novelty 6.0 of 10

    A curated 142-billion-point real-world multivariate time series corpus improves zero-shot forecasting when combined with existing synthetic and univariate pretraining data across four foundation models.

  4. Beyond Static Uncertainty: Modeling Temporal Uncertainty Dynamics for Probabilistic Time Series Forecasting

    cs.LG 2026-03 accept novelty 6.0 of 10

    A location-scale VAE with a GRU volatility path that transfers and evolves scale from look-back to horizon yields better CRPS/NMAE than strong probabilistic and point baselines on nine datasets.

  5. Diff-MN: Diffusion Parameterized MoE-NCDE for Continuous Time Series Generation with Irregular Observations

    cs.LG 2026-01 conditional novelty 6.0 of 10

    Diff-MN generates continuous, arbitrary-resolution time series from irregular observations by diffusing MoE-NCDE dynamics weights, reporting consistent wins over KO-VAE and GT-GAN on ten datasets.

  6. CTBench: Cryptocurrency Time Series Generation Benchmark

    q-fin.ST 2025-08 conditional novelty 6.0 of 10

    CTBench is the first crypto-focused time series generation benchmark, combining forecasting and statistical arbitrage tasks to rank eight generative models.

  7. Towards Time Series Generation Conditioned on Unstructured Natural Language

    cs.LG 2025-06 conditional novelty 6.0 of 10

    A diffusion model with BERT language conditioning can generate simple 100-step time series from natural language prompts, supported by a new 63,010-pair dataset.

  8. $K^2$VAE: A Koopman-Kalman Enhanced Variational AutoEncoder for Probabilistic Time Series Forecasting

    cs.LG 2025-05 conditional novelty 6.0 of 10

    Combining a learned Koopman linearization with a learned Kalman filter inside a VAE produces a probabilistic forecaster that beats existing methods on most tested short- and long-horizon datasets.

  9. Time Series Generation Under Data Scarcity: A Unified Generative Modeling Approach

    cs.LG 2025-05 conditional novelty 6.0 of 10

    A pre-trained diffusion model with dynamic channel adaptation and dataset tokens generates time series from a few examples and outperforms from-scratch baselines on a 12-dataset few-shot benchmark.

  10. T2S: High-resolution Time Series Generation with Text-to-Series Diffusion Models

    cs.LG 2025-05 conditional novelty 6.0 of 10

    T2S uses a length-adaptive VAE and flow-matching diffusion transformer to generate variable-length time series from text captions, trained on a new fragment-level caption dataset.

  11. TarDiff: Target-Oriented Diffusion Guidance for Synthetic Electronic Health Record Time Series Generation

    cs.LG 2025-04 conditional novelty 6.0 of 10

    TarDiff guides diffusion-based synthetic EHR generation with a gradient-alignment signal computed from a guidance set, reporting improved downstream mortality and ICU-stay classification versus prior generative models.

  12. HDT: Hierarchical Discrete Transformer for Multivariate Time Series Forecasting

    cs.LG 2025-02 conditional novelty 6.0 of 10

    HDT forecasts multivariate time series by generating a discrete coarse trend of the future, then generating finer target tokens conditioned on that predicted trend, outperforming prior methods on five datasets.

  13. TimeDP: Learning to Generate Multi-Domain Time Series with Domain Prompts

    cs.LG 2025-01 conditional novelty 6.0 of 10

    TimeDP conditions a diffusion model on prototype assignments extracted from a few example series, achieving strong in-domain and few-shot unseen-domain time series generation.

  14. Synthetic Data for Portfolios: A Throw of the Dice Will Never Abolish Chance

    q-fin.PM 2025-01 conditional novelty 6.0 of 10

    Generating excessive synthetic returns from small samples biases statistics, and generic GANs learn high-variance components that matter least for long-short portfolios.

  15. Population Aware Diffusion for Time Series Generation

    cs.LG 2025-01 conditional novelty 6.0 of 10

    PaD-TS trains time series diffusion models with a maximum mean discrepancy penalty on cross-correlation distributions, improving population-level property preservation.

  16. SynEnergy: Anomaly Semantic-Guided Diffusion for Synthetic Energy Data Generation

    cs.LG 2026-08 conditional novelty 5.5 of 10

    SynEnergy learns region-level anomaly semantics from residual energy data and injects them into a diffusion generator, improving anomaly preservation fidelity by an average of 12.21% over 11 baselines.

  17. Enhancing Irregular Time Series Forecasting with Continuous-Time Modeling Framework

    cs.LG 2026-07 conditional novelty 5.0 of 10

    WrapFlow combines continuous-time event/gap tokenization with simulation-free residual flow matching on a Transformer to improve irregular multivariate time-series forecasting.

  18. Parallel Complex Diffusion for Scalable Time Series Generation

    cs.LG 2026-02 conditional novelty 5.0 of 10

    PaCoDi generates time series by diffusing real and imaginary spectral components in parallel, cutting attention FLOPs roughly in half while improving benchmark scores.

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

    cs.LG 2025-08 conditional novelty 5.0 of 10

    DLGAN, a dual-layer GAN with a supervised sequence autoencoder and feature-space GAN, generates synthetic time series with stronger temporal dependencies than six baselines on four benchmark datasets.

  20. Stage-Diff: Stage-wise Long-Term Time Series Generation Based on Diffusion Models

    cs.LG 2025-08 conditional novelty 5.0 of 10

    Stage-Diff generates long multivariate time series in stages, decomposing each stage into multi-scale trends and using multi-channel convolution to carry information between stages.

  21. Physics-Informed Diffusion Models for Unsupervised Anomaly Detection in Multivariate Time Series

    cs.LG 2025-08 unverdicted novelty 5.0 of 10

    A weighted physics-informed loss schedule during diffusion training improves unsupervised anomaly detection in multivariate time series, according to the paper's experiments.

  22. Case Studies of Generative Machine Learning Models for Dynamical Systems

    eess.SY 2025-08 conditional novelty 5.0 of 10

    Physics-informed VAEs with Hamiltonian-based losses generate trajectories that match training distributions and satisfy optimal-control equations from as few as 200 to 500 samples.

  23. TabularARGN: A Flexible and Efficient Auto-Regressive Framework for Generating High-Fidelity Synthetic Data

    cs.LG 2025-01 conditional novelty 5.0 of 10

    TabularARGN is a lightweight any-order auto-regressive model that matches or beats state-of-the-art tabular synthetic data quality while training up to orders of magnitude faster, including for sequential two-table datasets.

  24. A GAN-Based Framework for Robust Data Synthesis in Satellite Internet Observations

    cs.AI 2026-06 conditional novelty 4.0 of 10

    On a two-day Starlink measurement subset, GT-GAN generates synthetic data that best preserves the real distribution under block-wise and point-wise missingness, outperforming SeriesGAN and a Temporal VAE at 40% missingness.

  25. HierCVAE: Hierarchical Attention-Driven Conditional Variational Autoencoders for Multi-Scale Temporal Modeling

    cs.LG 2025-08 reject novelty 4.0 of 10

    HierCVAE, an attention-CVAE hybrid, claims large gains in energy forecasting, but its reported 15-40% improvements are contradicted by its own tables.

Pith tools