- Rating
- 2
- Confidence
- 4
- Informativeness
- 1
- Presentation
- 3
- Soundness
- 3
Read review
Paper summary
DecoVAE decomposes the input sequence into trend, seasonal, and residual components within a clear temporal structure. For determining uncertainty, DecoVAE employs the residual components to measure the variance of the prediction. Specifically, DecoVAE calculates the residual by removing the trend and seasonal components from the original series. Through three parallel VAEs for handling three corresponding components, DecoVAE gets the prediction mean from trend and seasonal predictors, combined with the prediction variance from the volatility predictor via the Gaussian distribution. Experiments on seven datasets report that DecoVAE achieves the lowest mean CRPS on all datasets and trains faster / uses less memory than competing probabilistic models. The ablation study supports the efficient of DecoVAE‘s design.
Critical strengths
- DecoVAE differentiates between predictable structure in the mean and in uncertainty, situating this distinction within existing econometric and neural volatility research.
- Trend/season and volatility are implemented with distinct inductive components within three parallel VAEs, which is conceptually coherent.
- Experiments show that DecoVAE outperforms baselines in 7 real-world datasets, within speed/memory/model-size comparisons, which do matter in probabilistic forecasting in practice.
- The paper is well organized, with clear statements and logical explanations. The motivation, methodology, and experimental results are easy to follow, making the proposed framework accessible to readers.
Critical weaknesses
- Novelty is arguably incremental and not established at the level of the presentation. The proposed decomposition is a general method in time series forecasting. Decomposition of the original series into trend/seasonal components is well used in
Wu, H., Xu, J., Wang, J., & Long, M. (2021). Autoformer: Decomposition transformers with auto-correlation for long-term series forecasting. Advances in neural information processing systems, 34, 22419-22430.
Zhou, T., Ma, Z., Wen, Q., Wang, X., Sun, L., & Jin, R. (2022). Fedformer: Frequency enhanced decomposed transformer for long-term series forecasting. International conference on machine learning.
Treating the (de-trended, de-seasonalized) residual as the basis for time-varying prediction variance is usually used in
Kendall, A., & Gal, Y. (2017). What uncertainties do we need in bayesian deep learning for computer vision?. Advances in neural information processing systems, 30.
Shukla, S. N., & Marlin, B. (2021). Heteroscedastic temporal variational autoencoder for irregularly sampled time series. In International Conference on Learning Representations.
- As stated in line 467, DecoVAE does not capture the dependency across features. Thus, can I deem DecoVAE a channel-independent (CI) method? This may stem from the FFT on the detrended series, as stated in line 165, and should be clearly stated in the method description. Although the authors claim this design is for a lightweight model, but the channel dependency is critical for time series forecasting. The FFT can be employed in Zs in seasonal predictor like following work:
Yi, K., Fei, J., Zhang, Q., He, H., Hao, S., Lian, D., & Fan, W. (2024). Filternet: Harnessing frequency filters for time series forecasting. Advances in neural information processing systems, 37, 55115-55140.
- For experiments, the results are lacking in Weather, which is the standard dataset in probabilistic and deterministic time series forecasting task.
- The ablation study clarifies the effectiveness of DecoVAE’s design, but I am more interested in the selection of in line 168. Since this relates to the generation of the residual components, and the cutoff of frequency modes is critical to balance useful frequency modes and noise.
Questions for authors
See weakness
3 comments
Q2. As stated in line 467, DecoVAE does not capture the dependency across features. Thus, can I deem DecoVAE a channel-independent (CI) method? This may stem from the FFT on the detrended series, as stated in line 165, and should be clearly stated in the method description. Although the authors claim this design is for a lightweight model, but the channel dependency is critical for time series forecasting. The FFT can be employed in Zs in seasonal predictor like following work: Yi, K., Fei, J., Zhang, Q., He, H., Hao, S., Lian, D., & Fan, W. (2024). Filternet: Harnessing frequency filters for time series forecasting. Advances in neural information processing systems, 37, 55115-55140.
A2. Thank you for raising this point. We agree that our wording in line 467 is ambiguous. The phrase “does not explicitly capture structured cross-channel dependencies” can be interpreted as implying that DecoVAE is channel-independent, which is not what we intended.
DecoVAE is not channel-independent. All input channels are jointly processed by the component encoders, which map the full multivariate feature vector into shared latent representations. Thus, cross-channel information can be captured at the representation level. In the seasonal stream, the FFT is applied only along the temporal dimension; the resulting -dimensional frequency representation is then jointly projected into the latent space, so the use of FFT does not imply channel independence.
What we intended to state in the limitation is that DecoVAE does not explicitly model structured cross-channel or cross-horizon dependence in the conditional predictive distribution. The output likelihood uses diagonal covariance, So cross-channel dependencies may be represented through the shared latent variables, but no explicit non-diagonal predictive covariance is parameterized.
FilterNet [1] could provide richer frequency-domain dependency modeling at the representation level, but it would not by itself remove this probabilistic limitation. We will revise the manuscript to clarify that DecoVAE captures cross-channel information through multivariate latent representations, while its limitation is the lack of explicit cross-channel or cross-horizon covariance in the predictive distribution.
References
[1] K. Yi, J. Fei, Q. Zhang, H. He, S. Hao, D. Lian, and W. Fan. FilterNet: Harnessing Frequency Filters for Time Series Forecasting. NeurIPS, 2024.
Q3. For experiments, the results are lacking in Weather, which is the standard dataset in probabilistic and deterministic time series forecasting task.
A3. Thank you for pointing this out. We have added experiments on the Weather benchmark for both short-term () and long-term () forecasting. The results are summarized below.
For , DecoVAE achieves the best performance in both metrics, with CRPS 0.286 ± 0.003 and NMAE 0.316 ± 0.001. For , DecoVAE achieves the best NMAE (0.459 ± 0.005) and the second-best CRPS (0.431 ± 0.002), behind TSDiff. Overall, DecoVAE remains highly competitive on this standard forecasting benchmark across both short and long horizons.
| Dataset | Metric | DecoVAE (ours) | CLaST | VAE | LaST | DeepAR | TFM | TSDiff | DeNOTS |
|---|---|---|---|---|---|---|---|---|---|
| Weather () | CRPS | 0.286 ± 0.003 | 0.287 ± 0.004 | 0.683 ± 0.011 | 0.538 ± 0.057 | 0.360 ± 0.025 | 1.008 ± 0.322 | 0.327 ± 0.060 | 0.434 ± 0.022 |
| NMAE | 0.316 ± 0.001 | 0.319 ± 0.004 | 0.491 ± 0.014 | 0.676 ± 0.090 | 0.425 ± 0.017 | 1.206 ± 0.403 | 0.412 ± 0.093 | 0.562 ± 0.030 | |
| Weather () | CRPS | 0.431 ± 0.002 | 0.447 ± 0.001 | 0.661 ± 0.007 | 0.672 ± 0.028 | 0.465 ± 0.064 | 1.344 ± 0.246 | 0.375 ± 0.030 | 0.803 ± 0.071 |
| NMAE | 0.459 ± 0.005 | 0.482 ± 0.001 | 0.606 ± 0.008 | 0.835 ± 0.043 | 0.555 ± 0.067 | 1.542 ± 0.422 | 0.465 ± 0.033 | 0.865 ± 0.084 |
Bold denotes the best result and italics the second-best result; lower is better.
Q4. The ablation study clarifies the effectiveness of DecoVAE’s design, but I am more interested in the selection of in line 168. Since this relates to the generation of the residual components, and the cutoff of frequency modes is critical to balance useful frequency modes and noise.
A4. Thank you for raising this point. We treated as a dataset-specific hyperparameter controlling the fraction of spectral energy retained when selecting the seasonal frequency modes. It was selected jointly with the other hyperparameters using Optuna over .
Since directly controls the separation between the seasonal and residual components, we additionally evaluated the sensitivity of DecoVAE to this parameter. We used the same challenging synthetic setting introduced in our robustness analysis of the nonzero residual conditional mean, as it contains substantial time-varying residual dynamics and therefore provides a suitable stress test for the frequency cutoff. The corresponding quantitative results and decomposition visualizations are provided in https://anonymous.4open.science/r/DecoVAE-3387/rebuttal/vGnd_W4.pdf.
The results show that performance is worse at the boundary values, and , while intermediate values provide the best forecasting quality. This behavior is consistent with the intended role of the cutoff: when is too low, too few frequency modes are retained and part of the useful periodic structure remains in the residual component. Conversely, when is too high, additional noisy modes are absorbed into the seasonal representation instead of remaining in the residual.
Thus, intermediate values provide a better balance between retaining informative seasonal modes and separating residual uncertainty.
Q1. Novelty is arguably incremental and not established at the level of the presentation. The proposed decomposition is a general method in time series forecasting. Decomposition of the original series into trend/seasonal components is well used in Wu, H., Xu, J., Wang, J., & Long, M. (2021). Autoformer: Decomposition transformers with auto-correlation for long-term series forecasting. Advances in neural information processing systems, 34, 22419-22430.
Zhou, T., Ma, Z., Wen, Q., Wang, X., Sun, L., & Jin, R. (2022). Fedformer: Frequency enhanced decomposed transformer for long-term series forecasting. International conference on machine learning.
Treating the (de-trended, de-seasonalized) residual as the basis for time-varying prediction variance is usually used in
Kendall, A., & Gal, Y. (2017). What uncertainties do we need in bayesian deep learning for computer vision?. Advances in neural information processing systems, 30.
Shukla, S. N., & Marlin, B. (2021). Heteroscedastic temporal variational autoencoder for irregularly sampled time series. In International Conference on Learning Representations.
A1: Thank you for raising this concern. We agree that neither trend--seasonality decomposition nor heteroscedastic uncertainty modeling is novel in isolation. The contribution of DecoVAE is their role-specialized probabilistic organization: trend and seasonality parameterize the forecast mean, while a dedicated residual-volatility representation models time-varying forecast variance.
Autoformer [1] and FEDformer [2] use decomposition primarily to represent and predict temporal dynamics, but do not assign the residual a dedicated probabilistic role. Kendall & Gal [3] provide foundational work on input-dependent heteroscedastic uncertainty, although in a computer-vision setting rather than time-series forecasting. HeTVAE [4] is closer in the temporal domain, combining variational modeling with heteroscedastic outputs for irregular time series.
As summarized above, DecoVAE explicitly combines Trend--Seasonality--Residual decomposition with separate statistical roles: structured components model the forecast mean, while residual history is represented specifically to model future volatility.
Our new controlled ablation (https://anonymous.4open.science/r/DecoVAE-3387/rebuttal/ResidualModeling.pdf) further clarifies this contribution. On ETTh2, introducing the structured deterministic residual representation improves CRPS from 0.444 to 0.308, while stochastic sampling and KL further improve it to 0.293. Thus, the main gain comes from structured residual-volatility modeling, with variational modeling providing an additional, dataset-dependent generalization benefit.
We will revise the manuscript to make this distinction explicit.
References
[1] H. Wu, J. Xu, J. Wang, and M. Long. Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecasting. NeurIPS, 2021.
[2] T. Zhou, Z. Ma, Q. Wen, X. Wang, L. Sun, and R. Jin. FEDformer: Frequency Enhanced Decomposed Transformer for Long-term Series Forecasting. ICML, 2022.
[3] A. Kendall and Y. Gal. What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision? NeurIPS, 2017.
[4] S. N. Shukla and B. M. Marlin. Heteroscedastic Temporal Variational Autoencoder for Irregularly Sampled Time Series. ICLR, 2022.