- Rating
- 2
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
XMatch augments covariate-aware forecasting by retrieving historical endogenous response patterns using future exogenous inputs. Soft-DTW clustering and information-gain ordering organize training-set associations into a ProtoTree. A matcher distributes weights across branches and depths according to similarity and historical support, and fuses response prototypes with a Transformer representation. Experiments cover 12 datasets and 10 baselines; averaged across two forecasting settings, XMatch reports the best MSE on nine datasets and MAE on all twelve.
Critical strengths
- The paper observes that exogenous patterns often co-occur with a small set of endogenous response patterns in real data, highlighting reusable Exo-Endo correspondences that conventional covariate-aware forecasting leaves underexplored. The problem is clearly defined, and Exo-Endo Association Augmentation follows naturally from this observation.
- XMatch uses a ProtoTree Creator to organize historical associations and a ProtoTree Matcher for adaptive matching, yielding an interpretable design. Results on 12 datasets against 10 baselines, together with ablations and parameter analyses, provide a reasonably comprehensive experimental framework.
Critical weaknesses
- The method appears to combine existing retrieval, prototype memory, Soft-DTW clustering and attention routing techniques, without sufficiently establishing a substantive advance over prior retrieval/prototype methods. Adaptive matching depth also relies primarily on heuristic thresholds, with limited theoretical or formal analysis.
- Experiments concentrate on electricity, energy and wind-power scenarios with few exogenous variables and known future covariates, potentially favoring the proposed strategy. Analyses of tree depth, node support, matching depth, variable-ordering stability, clustering quality and computational overhead remain insufficient to explain the performance gains.
Questions for authors
- The method does not appear inherently tied to a particular Transformer. Could it use alternative backbones such as PatchTST, iTransformer, DLinear or Mamba to demonstrate the generality of the ProtoTree Matcher?
- Could the current reliance on patch shapes and Soft-DTW clustering be extended with frequency-domain or wavelet features, calendar information, conditional dependencies among variables, or uncertainty in exogenous forecasts? Such extensions may broaden applicability and mitigate the limitations of fixed variable ordering and patch length.
10 comments
Response to Reviewer tLqe (Part II)
W1-b: Justification for adaptive depth
We thank the reviewer for the opportunity to clarify the motivation for adaptive matching depth. As discussed in the Introduction, matching more exogenous variables provides more specific historical references, but leaves fewer samples to support them; matching fewer variables retains more evidence but may overlook relevant conditions. Adaptive depth allows XMatch to adjust this balance for each query and patch, using more specific matches when the evidence permits and retaining broader historical support otherwise.
The two gates implement this design through complementary checks: the similarity gate checks whether a candidate branch matches the query's exogenous pattern, while the support gate checks whether it has sufficient historical samples. Weight proceeds to deeper nodes only when both checks pass; otherwise, it remains at a shallower node. Together, these gates select relevant and sufficiently supported historical evidence to assist forecasting. Our contribution lies in this evidence-selection perspective and its realization through adaptive tree matching; the thresholds are practical controls selected on validation data. The newly added separate gate ablations (c1 and c2; Table 2, Lines 403–414; Lines 417–423) support this design: removing either gate degrades forecasting performance across the evaluated datasets.
| Dataset | BE | DE | Sdwpfh1 | Sdwpfh2 | Energy | |||||
|---|---|---|---|---|---|---|---|---|---|---|
| Metrics | MSE | MAE | MSE | MAE | MSE | MAE | MSE | MAE | MSE | MAE |
| (c) w/o search stopping | 0.430 | 0.283 | 0.380 | 0.375 | 0.419 | 0.454 | 0.447 | 0.484 | 0.099 | 0.246 |
| (c1) w/o similarity gate | 0.421 | 0.276 | 0.377 | 0.374 | 0.422 | 0.454 | 0.437 | 0.483 | 0.097 | 0.238 |
| (c2) w/o support gate | 0.423 | 0.276 | 0.374 | 0.373 | 0.457 | 0.462 | 0.449 | 0.491 | 0.096 | 0.239 |
| XMatch (Full) | 0.420 | 0.273 | 0.349 | 0.369 | 0.406 | 0.442 | 0.413 | 0.451 | 0.089 | 0.231 |
W2-a: Few exogenous variables may favor the method
The results do not suggest that XMatch is favored by having few exogenous variables: relative to the strongest baseline, its average MSE/MAE reductions are 10.0%/7.3% on the five datasets with at least five exogenous variables (Energy, Sdwpfh1, Sdwpfh2, Sdwpfm1 and Sdwpfm2), compared with 0.3%/2.7% on the seven datasets with two (NP, PJM, BE, FR, DE, Colbun and Rapel). This is consistent with adaptive matching being more useful when more exogenous combinations make the precision–support trade-off harder. These 12 datasets are also standard benchmarks used by DAG, KITE and GCGNet. Their widespread adoption in covariate-aware forecasting research also provides indirect support for their relevance to real-world forecasting scenarios.
Response to Reviewer tLqe (Part III)
We thank the reviewer for highlighting this important practical concern. In real-world settings, future covariates may be unavailable or affected by noise and disturbances, and handling these situations is essential. To address the case of unavailable future covariates, we have added an evaluation of the setting without future exogenous inputs in Appendix C.2 (Table 9, Lines 1247–1261; Table 10, Lines 1264–1294), where future covariates are replaced by mask tokens. XMatch still ranks first on 11/12 datasets in both MSE and MAE. This shows that the gains do not depend only on having future covariates. The table below reports mean MSE over two horizons for XMatch and four baselines; full MSE/MAE results are in Appendix C.2.
| Dataset (MSE) | XMatch | DAG | KITE | GCGNet | TimeXer |
|---|---|---|---|---|---|
| NP | 0.405 | 0.419 | 0.405 | 0.430 | 0.460 |
| PJM | 0.123 | 0.126 | 0.127 | 0.131 | 0.142 |
| BE | 0.452 | 0.462 | 0.453 | 0.459 | 0.501 |
| FR | 0.428 | 0.435 | 0.429 | 0.435 | 0.465 |
| DE | 0.597 | 0.603 | 0.601 | 0.614 | 0.683 |
| Energy | 0.160 | 0.150 | 0.146 | 0.154 | 0.174 |
| Sdwpfm1 | 0.686 | 0.689 | 0.698 | 0.711 | 0.744 |
| Sdwpfm2 | 0.781 | 0.786 | 0.800 | 0.807 | 0.867 |
| Sdwpfh1 | 0.724 | 0.733 | 0.760 | 0.741 | 0.784 |
| Sdwpfh2 | 0.860 | 0.870 | 0.880 | 0.886 | 0.925 |
| Colbun | 0.112 | 0.117 | 0.136 | 0.119 | 0.166 |
| Rapel | 0.238 | 0.261 | 0.302 | 0.265 | 0.380 |
| 1st Count | 11 | 0 | 2 | 0 | 0 |
Following Reviewer bXsN's comment, we found that the TimeXer results in this setting, which we had taken from DAG [1], contained errors. We have rerun TimeXer with the public DAG code and replaced these entries in Tables 9 and 10 with the corrected results, as shown above. XMatch's ranking and the conclusions are unchanged.
[1] DAG: A Dual Correlation Network for Time Series Forecasting with Exogenous Variables. ICML 2026.
Response to Reviewer tLqe (Part I)
We are encouraged to hear that Reviewer tLqe finds our observation of recurring exogenous–endogenous associations to be well motivated and clearly defined, and considers the ProtoTree-based design to be interpretable and supported by a reasonably comprehensive experimental framework. We hope the following discussions can address the reviewer’s remaining questions.
W1-a: Novelty beyond existing retrieval/prototype methods
Our contribution starts from the observation that recurring exogenous–endogenous associations can provide explicit historical evidence for forecasting. Exploiting these associations requires addressing two connected challenges: organizing them into a usable memory and balancing matching specificity against sufficient historical support when multiple exogenous variables are involved. We design XMatch as a unified framework which addresses the precision–support dilemma through a tree-structured association memory and adaptive matching conditioned on future exogenous patterns. Clustering, prototype construction, retrieval, and routing serve complementary roles in realizing this framework. Our main contribution therefore lies in how these components are jointly designed to support Exo–Endo Association Augmentation, with further methodological refinements within individual modules, as detailed below:
To further isolate the contribution of the tree structure, we followed the suggestions of Reviewers bXsN and MJsD and added a fairer, controlled flat-retrieval baseline. It uses the same prototypes, encoders, response mapping and backbone. It only replaces hierarchical routing and stopping with top-k scoring over all non-root nodes, with k and the temperature tuned on validation data (Table 2, Lines 403–414; Lines 427–431).