- Rating
- 2
- Confidence
- 3
- Informativeness
- 1
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
This paper presents Mixtures of Localized Diffusion Processes (MeLD), a framework for time-series forecasting that separates history-resolved structure from unresolved stochastic variation. The method employs a learnable localized time-frequency decomposition, context-dependent routing, and selective residual diffusion. Motivated by the Mori–Zwanzig formulation of partially observed dynamical systems, the authors aim to focus stochastic modeling on uncertain components rather than applying diffusion to the entire forecast.
Experiments on standard long-term forecasting benchmarks report competitive improvements over deterministic and diffusion-based baselines.
Critical strengths
-
Separating history-supported predictive structure from residual stochastic variation provides a reasonable motivation for selectively allocating diffusion modeling capacity.
-
The method achieves favorable MSE, MAE, and CRPS results across standard forecasting benchmarks. The deterministic evaluations cover multiple prediction horizons, and the authors provide additional controlled baseline comparisons and statistical analyses.
Critical weaknesses
-
The paper motivates its resolved/unresolved decomposition through Mori–Zwanzig theory, but does not adequately position this formulation relative to existing stochastic dynamical and state-space approaches. For instance, LLapDiff [1] already introduces a stochastic port-Hamiltonian formulation and locally linearized latent dynamics, separating structured mean evolution from stochastic excitation while employing history-conditioned modal representations. Although MeLD differs by introducing learned time-frequency routing and selective residual diffusion, the paper does not clearly establish how its proposed decomposition advances beyond these existing dynamical principles.
-
The paper defines conditional unresolved energy through the variance remaining after conditioning on the observed history. However, its implemented component-selection score is the product of learned routing weights and the temporal variance of the deterministic anchor. The relationship between this heuristic score and the formally defined conditional unresolved energy is not established. A component with high deterministic variation need not exhibit high residual uncertainty, and vice versa.
-
The paper motivates forecasting under partial observation and incomplete measurements, yet its main evaluations use standard regularly sampled forecasting benchmarks without explicitly controlling the degree of partial observation. While incomplete observation of an underlying dynamical state is not equivalent to missing entries in a measured time series, the current experiments do not directly examine how the method behaves as relevant state information becomes unavailable. The synthetic experiment validates recovery of deterministic and stochastic frequency bands, but does not systematically study different observation operators or levels of partial observability.
-
Although the proposed framework emphasizes localized unresolved dynamics, the implemented forecast-side decomposition uses a single forecast-window segment, with diffusion selection primarily operating over frequency and channel components. Consequently, the extent to which the method adaptively localizes stochastic variation across different physical-time regions of the forecast remains unclear. Furthermore, while the deterministic comparisons cover horizons of 96, 192, 336, and 720, the main comparisons against stochastic baselines are restricted to a single prediction horizon.
-
The main stochastic comparisons rely on previously published baseline results rather than a unified reproduction protocol, making it difficult to fully assess differences arising from sampling and evaluation settings. Moreover, CRPS and minimum-over-N error do not provide a complete assessment of the joint temporal dependence of generated trajectories. Since the proposed method explicitly aims to model structured stochastic variation, additional joint probabilistic metrics, such as energy score and variogram score, together with calibration diagnostics, would provide stronger evidence for the quality of the learned forecast distribution. The statistical comparisons against published baseline point estimates also do not account for uncertainty in those baseline results.
[1] You, Z., Zheng, J. & Cartlidge, J.. (2026). Latent Laplace Diffusion for Irregular Multivariate Time Series. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:149909-149935.
Questions for authors
-
On Novelty: How does MeLD's resolved/unresolved decomposition claim differ fundamentally from LLapDiff's Port Hamiltonian-SDE formulation and related structured state-space methods?
-
Routing justification: What theoretical or empirical evidence supports using anchor variance and routing weights as a proxy for conditional unresolved uncertainty?
-
Empirical validation: Can the authors evaluate the robustness of selective diffusion under controlled partial observability and across different probabilistic forecasting horizons?
7 comments
(W3) The evaluations do not control the degree of partial observation, and the synthetic experiment does not study different observation operators or levels of observability.
To evaluate selective diffusion under controlled partial observability, we add an experiment in which the number of observed stations is varied.
To this end, we generated synthetic time series by combining a large predictable sinusoid (period 96) with hidden signals and white noise. The hidden signal was defined as s(t) = ρ s(t − 96) + (1 − ρ²)^{1/2} e(t) with an innovation e that was never observed. This way, time-series forecast window has the unresolved fraction of 1 − (ρ² + ρ⁴)/2. We found that, across the unresolved fractions (degrees of partial observation), MeLD has shown the strongest performance compared to the alternatives. The gain of MeLD over the anchor grows with the unresolved fraction, from 0.036 CRPS at ρ = 0.95 to 0.153 at ρ = 0.30, because the less the history determines, the more of the forecast error is unresolved variation that only a generative model can represent, and MeLD places that model on the components that carry it, which is why its agreement with the oracle rises from 48% to 90% at the same time.
| ρ | 0.95 | 0.80 | 0.60 | 0.30 |
|---|---|---|---|---|
| Unresolved fraction | 0.14 | 0.48 | 0.76 | 0.95 |
| Anchor only | 0.220 | 0.387 | 0.490 | 0.561 |
| Oracle band (κ=1) | 0.188 | 0.288 | 0.360 | 0.413 |
| All bands (κ=5) | 0.244 | 0.315 | 0.369 | 0.412 |
| MeLD, selected κ | 0.184 (κ=1) | 0.290 (κ=1) | 0.360 (κ=1) | 0.408 (κ=2) |
| Top-1 = oracle | 48% | 69% | 78% | 90% |
(W4) Stochastic variation is not localized in time within the forecast window, and only one horizon is used for stochastic comparisons.
To evaluate the role of localized variation across physical time, we referred to the Ablations table (Table 5) of the submitted manuscript. The ablation “w/o forecast seg” compares the forecast made on localized vs. global time segments. We observed that, across 6 considered datasets, the use of localized time segments has consistently led to better forecasting.
To extend the probabilistic evaluation across the forecast horizons, we trained MeLD on horizons of 96, 192, 336, and 720. In each case, to evaluate the contribution of stochastic modeling, we evaluated the CRPS of the (deterministic) anchor forecast and (deterministic + stochastic) full MeLD forecast. We found that, in each considered case, full MeLD results were better than the anchor results, suggesting that our proposed method is efficient across forecasting horizons. We have additionally compared the results of our model, across horizons, to TMDM and found that our model outperforms it across all the horizons, further strengthening our results.
| Dataset | Model | H=96 | H=192 | H=336 | H=720 |
|---|---|---|---|---|---|
| Exchange | TMDM | 0.200 | 0.306 | 0.414 | 0.714 |
| Anchor | 0.198 | 0.293 | 0.403 | 0.792 | |
| MeLD (κ=1) | 0.158 | 0.245 | 0.349 | 0.690 | |
| ETTm2 | TMDM | 0.264 | 0.294 | 0.355 | 0.486 |
| Anchor | 0.253 | 0.293 | 0.334 | 0.398 | |
| MeLD (κ=1) | 0.200 | 0.234 | 0.268 | 0.324 | |
| Weather | TMDM | 0.202 | 0.239 | 0.275 | 0.333 |
| Anchor | 0.226 | 0.258 | 0.295 | 0.344 | |
| MeLD (κ=1) | 0.187 | 0.217 | 0.252 | 0.298 |
(W5) Limitations in probabilistic evaluation and baseline compatibility.
The main stochastic comparisons rely on previously published baseline results rather than a unified reproduction protocol, making it difficult to fully assess differences arising from sampling and evaluation settings.
To address this point, we’ve implemented a unified evaluation protocol that is the same for MeLD and the baselines. The published baseline numbers in Table 2 use the same datasets, splits, normalization, context length and horizon as our runs (table below).
CRPS and minimum-over-N error do not provide a complete assessment of the joint temporal dependence of generated trajectories. Since the proposed method explicitly aims to model structured stochastic variation, additional joint probabilistic metrics, such as energy score and variogram score. Together with calibration diagnostics, would provide stronger evidence for the quality of the learned forecast distribution.
To address this point, we’ve evaluated the suggested metrics (table below). We found that MeLD consistently shows the best performance across the requested metrics. We thank the Reviewer for this suggestion that has further strengthened our results.
| Dataset | Model | CRPS | ES | VS | Cov. 50% | Cov. 90% |
|---|---|---|---|---|---|---|
| Exchange | TMDM | 0.306 | 15.40 | 125.2 | 0.36 | 0.73 |
| Anchor only | 0.293 | 15.49 | 53.2 | – | – | |
| MeLD (κ=1) | 0.229 | 11.48 | 118.6 | 0.51 | 0.80 | |
| ETTm2 | TMDM (re-run) | 0.294 | 15.22 | 119.1 | 0.43 | 0.73 |
| Anchor only | 0.296 | 16.00 | 79.9 | – | – | |
| MeLD (κ=1) | 0.236 | 12.18 | 102.9 | 0.48 | 0.76 | |
| Weather | TMDM | 0.239 | 23.59 | 115.1 | 0.48 | 0.70 |
| Anchor only | 0.258 | 29.05 | 57.5 | – | – | |
| MeLD (κ=1) | 0.216 | 22.13 | 112.3 | 0.49 | 0.71 |
Response to Reviewer qYDp
(W1, Q1) Positioning, novelty and advances beyond existing stochastic dynamical and state-space methods, like LLapDiff [1]. How does MeLD's resolved/unresolved decomposition claim differ fundamentally from LLapDiff's Port Hamiltonian-SDE formulation and related structured state-space methods?
We appreciate the Reviewer bringing this method to our attention and pointing us to this line of work. The revision cites LLapDiff and the structured state-space methods in Related Work. Having studied LLapDiff closely, we find that it and MeLD differ as follows. Positioning. MeLD is a framework for deciding where stochastic modeling is needed in a forecast. Given the context window, it identifies the components of the forecast window that the context leaves unresolved; generative modeling is then only applied to those. LLapDiff, latent SDEs and deep state-space models address a different question, namely how a latent state evolves under noise, and they generate the whole state stochastically. The two questions are complementary. We evaluate MeLD on regularly sampled benchmarks, and LLapDiff is designed for and evaluated on irregularly sampled series.
Novelty. Compared to existing generative forecasting methods, as main contributions we view the selection mechanism and the localized generative processing based on its outputs. The selection step has no counterpart in LLapDiff or in latent SDE and state-space models. Beyond the selection mechanism, the formulations of MelD and LLapDiff share only the generic form of a mean term plus a noise term. In LLapDiff the noise term belongs to a single latent state and is always present. In MeLD it stands for the degrees of freedom that the history does not resolve, so it is tied to specific components and is absent from the others.
Advances over existing work. In our work we show that generative modeling of selected components of a time-series context window leads to better forecasts, compared to the modeling of the entire signal. For synthetic time series with a dominant predictable band, diffusion on all bands leads to a CRPS of 0.324. This number is worse than the result of the deterministic anchor alone (0.206), because generative modeling adds variation where the anchor is already correct. Diffusion on the selected band leads to the CRPS of 0.156, matching the oracle’s performance. On some real-world datasets (ring-transport, ETTm1) the selected band likewise leads to an improvement, compared to all-band diffusion, while on the others (Weather, ETTm2) it matches it, at about one fifth of the sampling time.
(W2, Q2) The selection score is not related to the conditional unresolved energy. What evidence supports anchor variance and routing weights as a proxy.
To respond to the Reviewer’s comment, we:
-Refer to the original submission’s table 4 that empirically shows positive Spearman rank correlations between the (routing weight times anchor variance) and conditional energy;
-Introduce Claim A.1 and Claim A.2 in Appendix section A.1 to show analytically that the Spearman rank correlations between the (routing weight times anchor variance) and conditional energy is positive;
Claim A.1 shows that the score, defined as routing weight times anchor variance, is positively rank-correlated with the residual energy, and gives the condition under which this holds.
Claim A.2 shows when the learned score is better than variance alone, when the router's weight increases with a component's noise-to-signal ratio.