- Rating
- 3(+1)
- Confidence
- 4
- Informativeness
- 2(+1)
- Presentation
- 3
- Soundness
- 2(+1)
Read review
Paper summary
This paper introduces HiGen, a hierarchical framework for generating irregular multivariate time series. By directly modeling the original observations through a Sample–Window–Point hierarchy, HiGen captures global characteristics, temporal dependencies across windows, and local interactions among asynchronously observed variables. To account for irregular sampling intervals, the framework further incorporates a learnable time-bin decoder. Extensive experiments on P12, Activity, and CesNet demonstrate promising results in downstream classification, generation fidelity, and the preservation of temporal irregularity, particularly in low-data regimes.
Critical strengths
-
Meaningful problem and well-motivated framework. The paper addresses the important challenge of irregular multivariate time series generation by directly modeling the original observations, thereby avoiding potential information loss associated with temporal alignment and interpolation.
-
Promising experimental results. The proposed Sample–Window–Point hierarchy offers a principled approach to capturing temporal dynamics and interactions among asynchronously observed variables. HiGen demonstrates strong empirical performance across three datasets, particularly in data-scarce settings.
Critical weaknesses
-
Serious concerns regarding reproducibility. Based on the authors' linked code repository, missing the
train.pyfile referenced in the README, along with three training entry points required by the low-data scripts. Furthermore, the released PRDC results (outputs/p12_lstm_ae_prdc.csv) report precision and recall values of approximately 0.9076 and 0.9107, respectively, which differ substantially from the corresponding values of 0.6430 and 0.6388 reported in Table 2, with no explanation provided for this discrepancy. Additional inconsistencies are evident in the experimental configurations: the checkpoint metadata specifiesnum_time_bins=8, whereas Appendix C states that K=10 is used for P12. Similarly, the low-data scripts default to training data proportions of 50%/70%/90%, rather than the 5%/10%/15% reported in the paper. Collectively, these discrepancies raise significant concerns about the reproducibility and reliability of the reported results, especially given the unusually large performance improvements. -
Insufficient experimental validation. The experimental evaluation lacks comparisons with recent time series generation methods, including Diffusion-TS (ICLR 2024) [1], L2D-Diff (ICLR 2026) [2], K-ProtoDiff (AAAI 2026) [3], and MA-TSD (ICML 2025) [4]. While some of these methods may require adaptation to handle irregularly sampled data, their omission limits the comprehensiveness of the evaluation. Moreover, the experiments are restricted to three relatively small datasets, with no systematic investigation of performance under varying sampling densities or degrees of temporal irregularity. The existing evaluation metrics also provide limited insight into the model's ability to preserve long-range temporal dependencies and cross-variable correlations. Additional metrics inspired by TimeGAN [5], such as discriminative and predictive scores, could offer complementary evidence if appropriately adapted to irregularly sampled data. Finally, the small font sizes in Figures 5 and 6 hinder readability.
[1] Yuan, X., & Qiao, Y. (2024). Diffusion-TS: Interpretable Diffusion for General Time Series Generation. ICLR 2024. [2] Shen, L., et al. (2026). Latent-to-Data Cascaded Diffusion Models for Unconditional Time Series Generation. ICLR 2026. [3] Duan, Y., Lin, L., & Wu, X. (2026). K-ProtoDiff: Key Prototypes-Guided Diffusion for Time Series Generation. AAAI 2026. [4] Wang, C., et al. (2025). A Non-isotropic Time Series Diffusion Model with Moving Average Transitions. ICML 2025. [5] Yoon, J., Jarrett, D., & van der Schaar, M. (2019). Time-series Generative Adversarial Networks. NeurIPS 2019.
Questions for authors
-
Based on the authors' linked code repository, several training scripts appear to be missing. Could the authors provide the missing files and complete reproduction instructions?
-
Could the authors clarify the discrepancies between the released PRDC results and Table 2, as well as the inconsistencies in
num_time_binsand low-data training proportions? -
Why were recent methods, such as Diffusion-TS (ICLR 2024), L2D-Diff (ICLR 2026), K-ProtoDiff (AAAI 2026), and MA-TSD (ICML 2025), excluded? Could the authors provide additional comparisons under appropriately adapted protocols?
-
Based on the authors' linked code repository, the evaluation covers only three datasets, two with fewer than 500 sequences. Could the authors include larger datasets or evaluate robustness across different sampling densities, missingness levels, and degrees of irregularity?
-
Could the authors further evaluate long-range temporal dependencies and cross-variable correlations, potentially using discriminative and predictive scores adapted from TimeGAN (NeurIPS 2019)?
4 comments
Response to Reviewer hLgp (Part Ⅱ)
Response to Q5: Long-Range Temporal Dependencies and Cross-Variable Correlations
Following your suggestion, we added Discriminative and Predictive Scores and a Lagged Value Correlation MAE on all three datasets. These evaluations examine overall sequence fidelity, temporal predictive relationships over different time gaps, and cross-variable correlations, respectively.
To evaluate overall sequence fidelity, the Discriminative Score measures how well an independent sequence classifier distinguishes real from synthetic sequences. Its inputs include observed values, masks, and actual time intervals. A lower score means that real and synthetic sequences are harder to distinguish.
To evaluate temporal dependencies over different forecast horizons, we train a multivariate predictor on synthetic data and test it on held-out real data. It uses past observations of all variables to predict future values at short, medium, and long horizons. The Predictive Score averages the prediction errors; lower values indicate better transfer of temporal predictive relationships.
To directly evaluate cross-variable correlations across time, we compare Pearson correlations between different variables at different time lags, without interpolation. The Lagged Value Correlation MAE measures the average absolute difference between real and synthetic correlations; lower values indicate closer recovery.
Table R18. Sequence Fidelity, Multi-Horizon Prediction, and Lagged Cross-Variable Correlations at r10.
| Dataset | Model | Discriminative Score ↓ | Predictive Score ↓ | Lagged Value Correlation MAE ↓ |
|---|---|---|---|---|
| P12 | HiGen | 0.2883 ± 0.0188 | 0.6006 ± 0.0010 | 0.1888 |
| P12 | TimEHR | 0.4929 ± 0.0007 | 0.5989 ± 0.0025 | 0.3510 |
| P12 | TimeDiff | 0.5000 ± 0.0000 | 2.8231 ± 0.0081 | 0.2236 |
| P12 | GT-GAN | 0.5000 ± 0.0000 | 0.6514 ± 0.0021 | 0.4983 |
| P12 | KoVAE | 0.4996 ± 0.0007 | 0.6411 ± 0.0043 | 0.3912 |
| Activity | HiGen | 0.3556 ± 0.0127 | 0.6228 ± 0.0019 | 0.2558 |
| Activity | TimEHR | 0.5000 ± 0.0000 | 0.7884 ± 0.0121 | 0.2896 |
| Activity | TimeDiff | 0.5000 ± 0.0000 | 0.8755 ± 0.0034 | 0.2076 |
| Activity | GT-GAN | 0.5000 ± 0.0000 | 0.6629 ± 0.0021 | 0.2039 |
| Activity | KoVAE | 0.5000 ± 0.0000 | 0.7135 ± 0.0035 | 0.3723 |
| CesNet | HiGen | 0.3750 ± 0.0278 | 0.6520 ± 0.0008 | 0.0752 |
| CesNet | TimEHR | 0.4907 ± 0.0080 | 0.8256 ± 0.0050 | 0.4692 |
| CesNet | TimeDiff | 0.5000 ± 0.0000 | 7.9308 ± 0.2294 | 0.0833 |
| CesNet | GT-GAN | 0.4491 ± 0.0882 | 0.7712 ± 0.0077 | 0.2498 |
| CesNet | KoVAE | 0.4907 ± 0.0080 | 0.7474 ± 0.0051 | 0.3393 |
Overall, HiGen performs favorably across these complementary metrics, supporting its ability to preserve temporal predictive relationships and lagged cross-variable correlations.
Response to W2: Readability of Figures 5 and 6
We increased the font sizes of axis labels, tick labels, and legends in Figures 5 and 6. We also adjust panel spacing and legend positions, and prioritize scalable vector versions. We check the figures at the normal reading size in the paper, rather than relying on enlarged screenshots. The numerical tables in this response also provide directly readable results and reduce reliance on small text in figures.
We hope our clarifications and additional experiments address your concerns. We would be happy to discuss any further questions!
Response to authors
Thank you for the further clarifications and additional results. I particularly appreciate the comparisons with recent baselines provided in your response, which I consider essential for a comprehensive evaluation of the work. Most of my concerns have been addressed, and I have increased my score accordingly.
Response to Reviewer hLgp
Thanks again for your insightful comments and positive feedback on our rebuttal. We truly appreciate the time and careful effort you put into reading our response and re-evaluating this work.
Response to Reviewer hLgp (Part Ⅰ)
Thank you for your thoughtful comments and constructive suggestions. They have helped us identify points that need clearer explanations and stronger experimental support. We address your concerns below and describe the corresponding revisions.
Response to Q1: Missing Training Files and Complete Reproduction Instructions
We organize the materials into a complete reproduction checklist. We provide the
train.pyfile referenced in the README and the training entry points required by the low-data scripts for the three datasets.Response to Q2/W1: Discrepancies in PRDC, the Number of Time Components, and Low-Data Training Proportions
Thank you for pointing out these inconsistencies.
PRDC naming. The precision/recall values of 0.9076/0.9107 come from hard-threshold PRDC. Table 2 reports 0.6430/0.6388 using a PRDC version based on continuous neighborhood affinity. These are different metrics. All PRDC results in the paper use the affinity-based version. The paper does not mix the two metrics. We have corrected the metric in the code repository.
K. We verified that the released P12 checkpoint uses K=8. The corresponding configuration in Appendix C should be changed to the actual value. This was a typo, and we have corrected it.
Low-data training proportions. The released scripts incorrectly used 50%/70%/90% as their default training proportions. These defaults were not used in the actual training runs. The experiments reported in the paper used 5%/10%/15%, corresponding to r5/r10/r15. We have corrected it.
Response to Q3: Comparisons with Recent Generative Methods
We added comparisons with Diffusion-TS, K-ProtoDiff, and MA-TSD after adapting them to our task. The table below shows synthetic-only LSTM results at r10. Each cell reports AUROC / AUPRC. Our model outperforms all these new baselines.
Table R16. Recent Generative Baselines at r10.
We have not obtained a verifiable official implementation of L2D-Diff, so we did not reproduce it in this round.
Response to Q4: Larger Datasets and Controlled Irregular Sampling
To address the concern about larger datasets, we added UEA LSST, which contains 4,925 sequences, 14 classes, and 6 variables. We used its official train/test split of 2,459/2,466 sequences.
To evaluate robustness across different sampling densities and missingness levels, we separately sampled each variable’s observations at the densities of 20%, 40%, 60%, and 80% as the training data, giving rise to 80%, 60%, 40%, and 20% missing rates.
Table R17. Synthetic-Only LSTM Results on LSST at r10.
HiGen has the highest mean AUROC/AUPRC under the listed conditions. The "—" entries for TimEHR mean that it could not generate valid synthetic samples because the input data were scarce. This experiment broadens the coverage of domains and classes while controlling sampling density.