Pith. sign in

REVIEW 4 major objections 6 minor 73 references

ST-FiT: Inductive Spatial-Temporal Forecasting with Limited Training Data

T0 review · 4 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read ST-FiT claims that any spatial-temporal GNN can be wrapped so it forecasts nodes with no temporal training data, without fine-tuning, by synthesizing diverse time series on a learned latent manifold and relearning sparse spatial topology.

desk verdict The problem setting is new and the method is sensible, but the headline gains over STGCN are not from a controlled comparison. read the letter →

arxiv 2412.10912 v2 pith:J35UZI6Z submitted 2024-12-14 cs.LG cs.AIstat.ML

Zhenyu Lei , Yushun Dong , Jundong Li , Chen Chen This is my paper · ORCID
classification cs.LGcs.AIstat.ML
keywords spatial-temporalforecastinginductivelearninglimitedtrainingdatatemporalaugmentationgraphtopologyvariationalautoencoderlatentmixupGumbel-Softmax
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

ST-FiT attacks a practical gap: most spatial-temporal forecasting models require every node to have a time series during training, but real deployments often have many nodes with no such history. The paper claims that any STGNN can be wrapped in a plug-and-play framework that forecasts for those cold-start nodes without fine-tuning. The framework learns a latent manifold of the available time series, interpolates between pairs of encodings to synthesize new series with varied temporal dependencies, and learns a sparse graph topology to support the richer training set. On three traffic datasets, ST-FiT beats all no-fine-tuning baselines and lands close to a fine-tuning-based competitor. If the claim holds, it broadens where spatial-temporal forecasting can be deployed, such as pandemic tracking and city-scale traffic prediction.

What carries the argument

The load-bearing object is a latent temporal manifold defined by a Variational Auto-Encoder over sliding windows of the available nodes' series. Temporal data augmentation performs mix-up in that latent space, $\hat z_v = \lambda z_{v_i} + (1-\lambda) z_{v_j}$ with $\lambda \in [0, 0.5]$, then decodes back to time series; this is meant to produce series that live close to the manifold while carrying new temporal dependencies. Spatial topology learning uses Gumbel-Softmax reparameterization over Bernoulli edge probabilities, with a sigmoid threshold that prunes spurious edges, to generate a sparse learned adjacency matrix used by the backbone. An iterative two-phase optimization alternates between training the augmentation module and jointly training the backbone with the topology learner, so the synthesized series and the graph structure are co-adapted.

What would settle it

Train ST-FiT on a small fraction of nodes from one traffic regime and evaluate it on nodes with sharply different periodicity or congestion patterns; if the VAE reconstruction error on those held-out nodes is much larger than on training nodes and forecasting gains disappear, the unified-manifold premise is violated.

Watch

Extended reading notes

Core claim

The central claim is that inductive forecasting with limited training data can be solved by enriching the training distribution rather than by adapting the test nodes. Given a spatial-temporal graph in which only a small subset of nodes has temporal data, ST-FiT uses a VAE to map available series into a latent space, assumes all nodes' series lie on one shared manifold, and creates new series by convex combinations of latent codes. A forecasting-aware similarity loss and a forecast loss push the synthesized series to carry consistent temporal dependencies. In parallel, a Gumbel-Softmax graph learner reweights and extends the adjacency matrix so the backbone sees spatial structure that fits both original and generated series. The two modules are optimized iteratively, and at inference the model simply runs the backbone on the learned topology. The paper reports that on PEMS03, PEMS04, and PEMS08 this recipe lets an STGCN backbone outperform FC-LSTM, STGODE, and STGCN without any fine-tuning, and match or beat the fine-tuned TransGTR baseline on most comparisons.

Load-bearing premise

All nodes' time series are assumed to sit on one shared latent manifold, so mixing the training nodes' encodings yields series that look like the unseen nodes' series; if the unseen nodes follow dynamics that are off that manifold, the synthetic training data misleads the backbone.

Editorial extensions

If this is right

  • Any STGNN backbone can be made inductive with no fine-tuning at inference time, so deployment on new nodes costs only one forward pass.
  • The stronger the temporal data limitation, the larger the reported gap over the no-fine-tuning baselines, with the biggest gains at 5% training nodes.
  • The method remains comparable to the backbone when all nodes have training data, so the augmentation and topology modules do not destroy standard forecasting performance.
  • The framework is evaluated on standard traffic datasets and metrics, making it straightforward to benchmark against future inductive forecasters.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A natural testable extension is to measure the VAE reconstruction error on held-out nodes before forecasting: if the unified-manifold assumption holds, that error should be no larger than for training nodes, and a large gap would warn that augmentation is producing off-manifold series.
  • The same recipe could be applied to epidemiology or energy grids, but the unified-manifold premise is more plausible when nodes are homogeneous; domain shifts such as different road types or climates would test its limits.
  • Because the topology learner is node-count agnostic and the backbone is frozen at inference, the framework could in principle be composed with larger pre-trained STGNNs, although the paper does not demonstrate that.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The paper introduces ST-FiT, a framework for inductive spatial-temporal forecasting with limited training data. The setting is that only a subset of nodes have temporal data during training, and the model must forecast for nodes with no temporal history. ST-FiT combines a VAE-based temporal data augmentation module, which mixes latent codes of training time series and decodes them into synthetic series, with a Gumbel-Softmax spatial topology learning module that refines the graph and generates edges for synthetic nodes. The framework is designed to wrap any STGNN backbone and is optimized in two alternating phases. Experiments on PEMS03, PEMS04, and PEMS08 report that ST-FiT outperforms non-fine-tuned baselines, including an STGCN baseline, by large margins (up to 40.0% MAE) and approaches the fine-tuned TransGTR baseline. Ablations, sensitivity analyses, and efficiency comparisons are included.

Significance. If the reported results hold, ST-FiT addresses a practically important and underexplored problem: generalizing an STGNN to nodes that have no temporal training data, without fine-tuning. The paper is clearly written, the code is released, and the main experiments use three standard datasets with three random seeds. The iterative optimization of augmentation and topology modules is a reasonable design, and the ablation study is comprehensive in its coverage of module variants. However, the central empirical claim rests on a comparison that is not fully controlled: the STGCN baseline and the STGCN backbone inside ST-FiT use different implementations and differ in whether layer normalization is retained. Because the ablations all use the modified backbone, they do not isolate the effect of the proposed modules alone. The claim that the framework works with 'any' STGNN is also not tested beyond STGCN.

major comments (4)
  1. [Experimental settings (Implementation of ST-FiT and Baselines), Table 1] The headline comparison against STGCN is not controlled. The STGCN baseline is run from the hazdzz/STGCN codebase, whereas the STGCN backbone inside ST-FiT follows the BasicTS implementation with layer normalization removed. Consequently, the reported 40.0% MAE improvement on PEMS08 (STGCN 41.67 vs. ST-FiT 25.09) conflates the effect of the proposed augmentation and topology modules with a different STGCN implementation and an architectural change. The ablation study in Table 2 does not repair this: every ablation (w/o aug, w/o gl, w/o sim, w/o fst, w/o gs, identity) uses the modified backbone, so there is no arm that measures the modified backbone alone with no augmentation and no learned topology. Please add this missing control arm and report its Table 1 metrics; without it, the main quantitative claim cannot be attributed to ST-FiT.
  2. [Equations (5), (9), and (10)] There is a sign inconsistency in the augmentation objective. The text states that the goal is to maximize the similarity between the generated code and the paired codes, and L_sim in Eq. (5) is written as a positive cosine similarity. However, L_sim is then added into Laug in Eq. (10), and Eq. (9) performs gradient descent on Laug. As written, minimizing Laug would minimize L_sim, pushing the generated code away from the pair, which contradicts the stated intent. Please either define L_sim as the negative cosine similarity or specify that Phase 1 maximizes L_sim; this is load-bearing for the augmentation module's behavior.
  3. [Abstract and Methodology (STGNN Backbone)] The abstract and Section 2 state that ST-FiT 'can be used on top of any existing STGNNs,' but all experiments use STGCN as the only backbone. The current evidence supports a claim about STGCN specifically, not about arbitrary STGNNs. To support the 'any backbone' claim, please evaluate at least one additional and structurally different backbone (e.g., a Graph WaveNet-style or transformer-based STGNN) under the same inductive protocol and with the same controlled comparison described above.
  4. [Temporal Data Augmentation, Eq. (3)] The paper explicitly assumes that latent codes from all nodes lie on a unified manifold and then acknowledges that the interpolation in Eq. (3) does not rigorously guarantee that the generated code stays on that manifold. Since this assumption is the main mechanism by which the augmentation creates useful training signal for unseen nodes, the paper should provide some direct evidence for it, for example by visualizing the latent geometry of training versus held-out test nodes or by showing that generated series are closer to test-node series than simple input-space baselines. Without such evidence, the generalization mechanism remains plausible but unverified.
minor comments (6)
  1. [Eq. (6)] In Eq. (6), the notation (·)[:ξ] and (·)[−τ:] is confusing: the generated sequence from Eq. (4) has length κ + τ, so the first ξ steps are not necessarily the κ input steps used by the backbone. Please clarify the intended slicing, or use the encoder input length κ consistently.
  2. [Algorithm 1] In Algorithm 1, line 17 reads 'Compute Laug following Eq. (12)', but Eq. (12) defines Lgf, not Laug. The algorithm should refer to Lgf in Phase 2 and to the correct update variable θgf rather than the undefined f.
  3. [Definition 1] In Definition 1, the text says 'Aij = 1 indicates that an edge exits between vi and vj'; the word should be 'exists'.
  4. [Efficiency Study] The efficiency study text says 'Table 8', but the referenced object is a figure (Figure 8). Please correct the cross-reference.
  5. [Packages Required for Implementations] The package list contains 'pandas == 2.2.2' twice. Please remove the duplicate entry.
  6. [Figure captions] Figures 6 and 7 describe the sparse threshold with the symbol ø, while the main text and Section 4.4 use ϵ. Please unify the notation.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: ST-FiT's reported generalization is evaluated on held-out nodes; the main caveat is a baseline confound, not a derivation that reduces to its inputs.

full rationale

ST-FiT makes an empirical claim (forecasting on nodes excluded from training), not a derivation. The reported metric is computed on held-out nodes whose labels never enter any training loss, so no predicted quantity is defined by a fitted parameter. The temporal augmentation module is trained to generate synthetic series and to be forecastable by the backbone (L_fst), but this only shapes the training signal; the headline comparison in Table 1 is an independent test on withheld nodes. The 'unified manifold' assumption is explicitly stated and even conceded to be unguaranteed, not imported from a self-citation. No self-citation chain is used to justify the central claim, and no existing result is renamed as a new prediction. The only substantive experimental caveat, that the STGCN comparison is not perfectly controlled because the ST-FiT backbone removes layer normalization and uses a different STGCN codebase than the STGCN baseline, concerns confounds in the empirical support rather than circularity in the sense of an equation reducing to its inputs. I therefore find no circular step and assign score 0.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The method introduces no new entities. Its central claim rests on two data-distribution assumptions (unified temporal manifold, feature-based topology transfer) and on several hand-chosen hyperparameters (λ, ε, φ, s, K). These are the main costs the reader pays beyond the standard STGNN backbone.

free parameters (5)
  • λ (mix-up ratio) = 0.5
    Interpolation weight in Eq. 3; selected for best performance in Figure 4/6. It controls the diversity of generated temporal data.
  • ε (sparse threshold) = 0.9
    Threshold in Eq. 8 controlling edge sparsity; selected for best performance in Figure 4/7. It controls how many edges are retained in the learned topology.
  • φ (sparsification temperature) = 0.1
    Temperature in Eq. 8; set by hand. Affects how sharply the threshold is applied.
  • s (Gumbel-Softmax temperature) = not reported
    Temperature in Eq. 7 for discrete topology sampling; left unspecified in the main text.
  • K (number of mixed pairs) = not reported
    Number of pairs sampled in each augmentation step (Eq. 3 context); not reported in the paper.
assumptions (4)
  • domain assumption All node time series lie on a unified manifold in the VAE latent space.
    Invoked before Eq. 2-3 ('We assume that z_v from all nodes come from a unified manifold in the hidden space'). If false, interpolating latent codes generates off-manifold data that does not help unseen nodes.
  • domain assumption The learned edge probability P derived from node features transfers to unseen nodes.
    Invoked in Spatial Topology Learning and at inference (Eq. 7 used on all nodes). Requires that feature-based similarity is sufficient to predict spatial dependencies for nodes not seen in training.
  • ad hoc to paper Forecasting error on generated data (L_fst) is a valid proxy for real temporal consistency.
    Used in Eq. 6 as a training objective; this assumes the backbone can serve as a consistency check for synthetic series.
  • standard math Standard VAE reparameterization and Gumbel-Softmax provide differentiable training signal.
    Used in Eq. 2 and Eq. 7; standard results from Kingma & Welling 2013 and Franceschi et al. 2019.

how reviews work

0 comments
Cite this review

Pith. "Pith review of ST-FiT: Inductive Spatial-Temporal Forecasting with Limited Training Data." pith.science (2026). https://pith.science/paper/J35UZI6Z

@misc{pith2026241210912,
  author       = {Pith},
  title        = {Pith review of: ST-FiT: Inductive Spatial-Temporal Forecasting with Limited Training Data},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/J35UZI6Z}},
  note         = {Machine review of arXiv:2412.10912}
}
read the original abstract

Spatial-temporal graphs are widely used in a variety of real-world applications. Spatial-Temporal Graph Neural Networks (STGNNs) have emerged as a powerful tool to extract meaningful insights from this data. However, in real-world applications, most nodes may not possess any available temporal data during training. For example, the pandemic dynamics of most cities on a geographical graph may not be available due to the asynchronous nature of outbreaks. Such a phenomenon disagrees with the training requirements of most existing spatial-temporal forecasting methods, which jeopardizes their effectiveness and thus blocks broader deployment. In this paper, we propose to formulate a novel problem of inductive forecasting with limited training data. In particular, given a spatial-temporal graph, we aim to learn a spatial-temporal forecasting model that can be easily generalized onto those nodes without any available temporal training data. To handle this problem, we propose a principled framework named ST-FiT. ST-FiT consists of two key learning components: temporal data augmentation and spatial graph topology learning. With such a design, ST-FiT can be used on top of any existing STGNNs to achieve superior performance on the nodes without training data. Extensive experiments verify the effectiveness of ST-FiT in multiple key perspectives.

Figures

Figures reproduced from arXiv: 2412.10912 by the authors.

Figure 1
Figure 1. An exemplary spatial-temporal graph where only [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. An overview of ST-FiT, including a STGNN backbone, temporal data augmentation, and spatial topology learning. [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. The performance of ST-FiT compared to baselines with different training node ratios. As training node ratio decreases, [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Performance of ST-FiT with different mix-up ra [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: The performance of ST-FiT compared to baselines with different training node ratios. As training node ratio decreases, [PITH_FULL_IMAGE:figures/full_fig_p011_5.png]
Figure 6
Figure 6. Figure 6: Performance of ST-FiT with different values of [PITH_FULL_IMAGE:figures/full_fig_p013_6.png]
Figure 7
Figure 7. Figure 7: Performance of ST-FiT with different sparse [PITH_FULL_IMAGE:figures/full_fig_p014_7.png]
Figure 8
Figure 8. Figure 8: Comparison of training time per epoch between [PITH_FULL_IMAGE:figures/full_fig_p014_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

73 extracted references · 47 canonical work pages

  1. [1]

    , " * write output.state after.block = add.period write newline

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.a...

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    N.; and Sokadjo, Y

    Atchad \'e , M. N.; and Sokadjo, Y. M. 2022. Overview and cross-validation of COVID-19 forecasting univariate models. Alexandria Engineering Journal, 61(4): 3021--3036

    work page 2022

  4. [4]

    Bai, L.; Yao, L.; Li, C.; Wang, X.; and Wang, C. 2020. Adaptive graph convolutional recurrent network for traffic forecasting. Advances in neural information processing systems, 33: 17804--17815

    2020

  5. [5]

    Bandara, K.; Hewamalage, H.; Liu, Y.-H.; Kang, Y.; and Bergmeir, C. 2021. Improving the accuracy of global forecasting models using time series data augmentation. Pattern Recognition, 120: 108148

    work page 2021

  6. [6]

    Cheng, Y.; Chen, P.; Guo, C.; Zhao, K.; Wen, Q.; Yang, B.; and Jensen, C. S. 2023. Weakly guided adaptation for robust time series forecasting. Proceedings of the VLDB Endowment, 17(4): 766--779

    work page 2023

  7. [7]

    Colan, S. D. 2013. The why and how of Z scores. Journal of the American Society of Echocardiography, 26(1): 38--40

    work page 2013

  8. [8]

    Cui, H.; Dai, W.; Zhu, Y.; Kan, X.; Gu, A. A. C.; Lukemire, J.; Zhan, L.; He, L.; Guo, Y.; and Yang, C. 2022. Braingb: a benchmark for brain network analysis with graph neural networks. IEEE transactions on medical imaging, 42(2): 493--506

    work page 2022

Show all 73 references
  1. [9]

    Cui, Y.; Zheng, K.; Cui, D.; Xie, J.; Deng, L.; Huang, F.; and Zhou, X. 2021. METRO: a generic graph neural network framework for multivariate time series forecasting. Proceedings of the VLDB Endowment, 15(2): 224--236

    2021

  2. [10]

    Dai, R.; Xu, S.; Gu, Q.; Ji, C.; and Liu, K. 2020. Hybrid spatio-temporal graph convolutional network: Improving traffic prediction with navigation data. In Proceedings of the 26th acm sigkdd international conference on knowledge discovery & data mining, 3074--3082

    2020

  3. [11]

    Deng, J.; Chen, X.; Jiang, R.; Song, X.; and Tsang, I. W. 2021. St-norm: Spatial and temporal normalization for multi-variate time series forecasting. In Proceedings of the 27th ACM SIGKDD conference on knowledge discovery & data mining, 269--278

    2021

  4. [12]

    Fang, Z.; Long, Q.; Song, G.; and Xie, K. 2021. Spatial-temporal graph ode networks for traffic flow forecasting. In Proceedings of the 27th ACM SIGKDD conference on knowledge discovery & data mining, 364--373

    2021

  5. [13]

    Fang, Z.; Wu, D.; Pan, L.; Chen, L.; and Gao, Y. 2022. When Transfer Learning Meets Cross-City Urban Flow Prediction: Spatio-Temporal Adaptation Matters. In IJCAI, volume 22, 2030--2036

    2022

  6. [14]

    Feng, A.; and Tassiulas, L. 2022. Adaptive graph spatial-temporal transformer network for traffic forecasting. In Proceedings of the 31st ACM international conference on information & knowledge management, 3933--3937

    2022

  7. [15]

    Franceschi, L.; Niepert, M.; Pontil, M.; and He, X. 2019. Learning discrete structures for graph neural networks. In International conference on machine learning, 1972--1982. PMLR

    2019

  8. [16]

    Fu, B.; Kirchbuchner, F.; and Kuijper, A. 2020. Data augmentation for time series: traditional vs generative models on capacitive proximity time series. In Proceedings of the 13th ACM international conference on pervasive technologies related to assistive environments, 1--10

    2020

  9. [17]

    A.; Jarvis, C

    Gibbs, H.; Liu, Y.; Pearson, C. A.; Jarvis, C. I.; Grundy, C.; Quilty, B. J.; Diamond, C.; and Eggo, R. M. 2020. Changing travel patterns in China during the early stages of the COVID-19 pandemic. Nature communications, 11(1): 5012

    2020

  10. [18]

    Goodfellow, I.; Pouget-Abadie, J.; Mirza, M.; Xu, B.; Warde-Farley, D.; Ozair, S.; Courville, A.; and Bengio, Y. 2020. Generative adversarial networks. Communications of the ACM, 63(11): 139--144

    2020

  11. [19]

    Goubeaud, M.; Jou en, P.; Gmyrek, N.; Ghorban, F.; Schelkes, L.; and Kummert, A. 2021. Using variational autoencoder to augment sparse time series datasets. In 2021 7th international conference on optimization and applications (ICOA), 1--6. IEEE

    2021

  12. [20]

    Guo, S.; Lin, Y.; Feng, N.; Song, C.; and Wan, H. 2019. Attention based spatial-temporal graph convolutional networks for traffic flow forecasting. In Proceedings of the AAAI conference on artificial intelligence, volume 33, 922--929

    2019

  13. [21]

    Guo, S.; Lin, Y.; Gong, L.; Wang, C.; Zhou, Z.; Shen, Z.; Huang, Y.; and Wan, H. 2023. Self-supervised spatial-temporal bottleneck attentive network for efficient long-term traffic forecasting. In 2023 IEEE 39th International Conference on Data Engineering (ICDE), 1585--1596. IEEE

    2023

  14. [22]

    Gupta, M.; Kodamana, H.; and Ranu, S. 2023. Frigate: Frugal spatio-temporal forecasting on road networks. In Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 649--660

    2023

  15. [23]

    Han, L.; Du, B.; Sun, L.; Fu, Y.; Lv, Y.; and Xiong, H. 2021. Dynamic and multi-faceted spatio-temporal deep learning for traffic speed forecasting. In Proceedings of the 27th ACM SIGKDD conference on knowledge discovery & data mining, 547--555

    2021

  16. [24]

    Hu, Z.; Zhao, Z.; Yi, X.; Yao, T.; Hong, L.; Sun, Y.; and Chi, E. 2022. Improving multi-task generalization via regularizing spurious correlation. Advances in Neural Information Processing Systems, 35: 11450--11466

    2022

  17. [25]

    M.; KIM, Y.; Kim, D.; et al

    Huh, I.; Choe, J. M.; KIM, Y.; Kim, D.; et al. 2024. Isometric Quotient Variational Auto-Encoders for Structure-Preserving Representation Learning. Advances in Neural Information Processing Systems, 36

    2024

  18. [26]

    K.; and Uchida, S

    Iwana, B. K.; and Uchida, S. 2021. Time series data augmentation for neural networks by time warping with a discriminative teacher. In 2020 25th International Conference on Pattern Recognition (ICPR), 3558--3565. IEEE

    2021

  19. [27]

    Jiang, J.; Wu, B.; Chen, L.; Zhang, K.; and Kim, S. 2023 a . Enhancing the Robustness via Adversarial Learning and Joint Spatial-Temporal Embeddings in Traffic Forecasting. In Proceedings of the 32nd ACM International Conference on Information and Knowledge Management, 987--996

    2023

  20. [28]

    Jiang, R.; Wang, Z.; Yong, J.; Jeph, P.; Chen, Q.; Kobayashi, Y.; Song, X.; Fukushima, S.; and Suzumura, T. 2023 b . Spatio-temporal meta-graph learning for traffic forecasting. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 37, 8078--8086

    2023

  21. [29]

    Jin, G.; Liang, Y.; Fang, Y.; Shao, Z.; Huang, J.; Zhang, J.; and Zheng, Y. 2023. Spatio-temporal graph neural networks for predictive learning in urban computing: A survey. IEEE Transactions on Knowledge and Data Engineering

    2023

  22. [30]

    Jin, W.; Ma, Y.; Liu, X.; Tang, X.; Wang, S.; and Tang, J. 2020. Graph structure learning for robust graph neural networks. In Proceedings of the 26th ACM SIGKDD international conference on knowledge discovery & data mining, 66--74

    2020

  23. [31]

    Jin, Y.; Chen, K.; and Yang, Q. 2023. Transferable Graph Structure Learning for Graph-based Traffic Forecasting Across Cities. In Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 1032--1043

    2023

  24. [32]

    Kefalas, P.; Symeonidis, P.; and Manolopoulos, Y. 2018. Recommendations based on a heterogeneous spatio-temporal social network. World Wide Web, 21: 345--371

    2018

  25. [33]

    P.; and Welling, M

    Kingma, D. P.; and Welling, M. 2013. Auto-encoding variational bayes. arXiv preprint arXiv:1312.6114

    2013 arXiv

  26. [34]

    R.; Sharma, Y.; Everett, K.; Priol, R

    Lachapelle, S.; L \'o pez, P. R.; Sharma, Y.; Everett, K.; Priol, R. L.; Lacoste, A.; and Lacoste-Julien, S. 2024. Nonparametric Partial Disentanglement via Mechanism Sparsity: Sparse Actions, Interventions and Sparse Temporal Dependencies. arXiv preprint arXiv:2401.04890

    2024 arXiv

  27. [35]

    Lan, S.; Ma, Y.; Huang, W.; Wang, W.; Yang, H.; and Li, P. 2022. Dstagnn: Dynamic spatial-temporal aware graph neural network for traffic flow forecasting. In International conference on machine learning, 11906--11917. PMLR

    2022

  28. [36]

    Li, F.; Yan, H.; Jin, G.; Liu, Y.; Li, Y.; and Jin, D. 2022. Automated spatio-temporal synchronous modeling with multiple graphs for traffic prediction. In Proceedings of the 31st ACM International Conference on Information & Knowledge Management, 1084--1093

    2022

  29. [37]

    Li, H.; Jin, D.; Li, X.; Huang, J.; Ma, X.; Cui, J.; Huang, D.; Qiao, S.; and Yoo, J. 2023. Dmgf-net: an efficient dynamic multi-graph fusion network for traffic prediction. ACM Transactions on Knowledge Discovery from Data, 17(7): 1--19

    2023

  30. [38]

    Li, M.; and Zhu, Z. 2021. Spatial-temporal fusion graph neural networks for traffic flow forecasting. In Proceedings of the AAAI conference on artificial intelligence, volume 35, 4189--4196

    2021

  31. [39]

    Li, T.; Zhang, J.; Bao, K.; Liang, Y.; Li, Y.; and Zheng, Y. 2020. Autost: Efficient neural architecture search for spatio-temporal prediction. In Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, 794--802

    2020

  32. [40]

    Liang, W.; Li, Y.; Xie, K.; Zhang, D.; Li, K.-C.; Souri, A.; and Li, K. 2022 a . Spatial-temporal aware inductive graph neural network for C-ITS data recovery. IEEE Transactions on Intelligent Transportation Systems

    2022

  33. [41]

    Liang, Y.; Shao, Z.; Wang, F.; Zhang, Z.; Sun, T.; and Xu, Y. 2022 b . BasicTS: An Open Source Fair Multivariate Time Series Prediction Benchmark. In International Symposium on Benchmarking, Measuring and Optimization, 87--101. Springer

    2022

  34. [42]

    Liu, H.; Dong, Z.; Jiang, R.; Deng, J.; Deng, J.; Chen, Q.; and Song, X. 2023. Spatio-temporal adaptive embedding makes vanilla transformer sota for traffic forecasting. In Proceedings of the 32nd ACM international conference on information and knowledge management, 4125--4129

    2023

  35. [43]

    Liu, Y.; Bahadori, T.; and Li, H. 2012. Sparse-gev: Sparse latent space model for multivariate extreme value time serie modeling. arXiv preprint arXiv:1206.4685

    2012 arXiv

  36. [44]

    Ouyang, C.; Chen, C.; Li, S.; Li, Z.; Qin, C.; Bai, W.; and Rueckert, D. 2022. Causality-inspired single-source domain generalization for medical image segmentation. IEEE Transactions on Medical Imaging, 42(4): 1095--1106

    2022

  37. [45]

    Ouyang, X.; Yang, Y.; Zhou, W.; Zhang, Y.; Wang, H.; and Huang, W. 2023. Citytrans: Domain-adversarial training with knowledge transfer for spatio-temporal prediction across cities. IEEE Transactions on Knowledge and Data Engineering

    2023

  38. [46]

    Panagopoulos, G.; Nikolentzos, G.; and Vazirgiannis, M. 2021. Transfer graph neural networks for pandemic forecasting. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 35, 4838--4845

    2021

  39. [47]

    M.; and Kim, Y

    Park, Y. M.; and Kim, Y. 2014. A spatially filtered multilevel model to account for spatial dependency: application to self-rated health status in South Korea. International journal of health geographics, 13: 1--10

    2014

  40. [48]

    Paszke, A.; Gross, S.; Chintala, S.; Chanan, G.; Yang, E.; DeVito, Z.; Lin, Z.; Desmaison, A.; Antiga, L.; and Lerer, A. 2017. Automatic differentiation in pytorch

    2017

  41. [49]

    PeMS, C. 2021. Caltrans performance measurement system (pems). US DOT

    2021

  42. [50]

    A.; and Awad, M

    Sahili, Z. A.; and Awad, M. 2023. Spatio-temporal graph neural networks: A survey. arXiv preprint arXiv:2301.10569

    2023 arXiv

  43. [51]

    Shang, C.; Chen, J.; and Bi, J. 2021. Discrete graph structure learning for forecasting multiple time series. arXiv preprint arXiv:2101.06861

    2021 arXiv

  44. [52]

    Shao, Z.; Wang, F.; Xu, Y.; Wei, W.; Yu, C.; Zhang, Z.; Yao, D.; Jin, G.; Cao, X.; Cong, G.; et al. 2023. Exploring progress in multivariate time series forecasting: Comprehensive benchmarking and heterogeneity analysis. arXiv preprint arXiv:2310.06119

    2023 arXiv

  45. [53]

    Shao, Z.; Zhang, Z.; Wang, F.; and Xu, Y. 2022. Pre-training enhanced spatial-temporal graph neural network for multivariate time series forecasting. In Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 1567--1577

    2022

  46. [54]

    Shin, Y.; and Yoon, Y. 2024. PGCN: Progressive graph convolutional networks for spatial--temporal traffic forecasting. IEEE Transactions on Intelligent Transportation Systems

    2024

  47. [55]

    Song, C.; Lin, Y.; Guo, S.; and Wan, H. 2020. Spatial-temporal synchronous graph convolutional networks: A new framework for spatial-temporal network data forecasting. In Proceedings of the AAAI conference on artificial intelligence, volume 34, 914--921

    2020

  48. [56]

    Sutskever, I.; and Vinyals, O. 2014. Sequence to sequence learning with neural networks. Advances in neural information processing systems

    2014

  49. [57]

    H.; Lam, W

    Tang, Y.; Qu, A.; Chow, A. H.; Lam, W. H.; Wong, S. C.; and Ma, W. 2022. Domain adversarial spatial-temporal network: A transferable framework for short-term traffic forecasting across cities. In Proceedings of the 31st ACM International Conference on Information & Knowledge M...

    2022

  50. [58]

    T.; Pfister, F

    Um, T. T.; Pfister, F. M.; Pichler, D.; Endo, S.; Lang, M.; Hirche, S.; Fietzek, U.; and Kuli \'c , D. 2017. Data augmentation of wearable sensor data for parkinson’s disease monitoring using convolutional neural networks. In Proceedings of the 19th ACM international conferenc...

    2017

  51. [59]

    Verma, V.; Lamb, A.; Beckham, C.; Najafi, A.; Mitliagkas, I.; Lopez-Paz, D.; and Bengio, Y. 2019. Manifold mixup: Better representations by interpolating hidden states. In International conference on machine learning, 6438--6447. PMLR

    2019

  52. [60]

    Wang, L.; Adiga, A.; Chen, J.; Sadilek, A.; Venkatramanan, S.; and Marathe, M. 2022. Causalgnn: Causal-based graph neural networks for spatio-temporal epidemic forecasting. In Proceedings of the AAAI conference on artificial intelligence, volume 36, 12191--12199

    2022

  53. [61]

    Wang, S.; Miao, H.; Li, J.; and Cao, J. 2021 a . Spatio-temporal knowledge transfer for urban crowd flow prediction via deep attentive adaptation networks. IEEE Transactions on Intelligent Transportation Systems, 23(5): 4695--4705

    2021

  54. [62]

    Wang, X.; Ma, Y.; Wang, Y.; Jin, W.; Wang, X.; Tang, J.; Jia, C.; and Yu, J. 2020. Traffic flow prediction via spatial temporal graph neural network. In Proceedings of the web conference 2020, 1082--1092

    2020

  55. [63]

    Wang, X.; Wang, H.; Wang, Z.; Lu, S.; and Fan, Y. 2021 b . Risk spillover network structure learning for correlated financial assets: A directed acyclic graph approach. Information Sciences, 580: 152--173

    2021

  56. [64]

    Wijsen, J. 2018. Temporal Dependencies

    2018

  57. [65]

    Wu, Y.; Yang, X.; Tang, Y.; Zhang, C.; Zhang, G.; and Zhang, W. 2022. Inductive spatiotemporal graph convolutional networks for short-term quantitative precipitation forecasting. IEEE Transactions on Geoscience and Remote Sensing, 60: 1--18

    2022

  58. [66]

    Wu, Z.; Pan, S.; Long, G.; Jiang, J.; and Zhang, C. 2019. Graph wavenet for deep spatial-temporal graph modeling. arXiv preprint arXiv:1906.00121

    2019 arXiv

  59. [67]

    Ye, J.; Zhao, J.; Ye, K.; and Xu, C. 2020. Multi-stgcnet: A graph convolution based spatial-temporal framework for subway passenger flow forecasting. In 2020 International joint conference on neural networks (IJCNN), 1--8. IEEE

    2020

  60. [68]

    Yu, B.; Yin, H.; and Zhu, Z. 2017. Spatio-temporal graph convolutional networks: A deep learning framework for traffic forecasting. arXiv preprint arXiv:1709.04875

    2017 arXiv

  61. [69]

    Zhang, C.; and Patras, P. 2018. Long-term mobile traffic forecasting using deep spatio-temporal neural networks. In Proceedings of the Eighteenth ACM International Symposium on Mobile Ad Hoc Networking and Computing, 231--240

    2018

  62. [70]

    Zhang, X.; Huang, C.; Xu, Y.; Xia, L.; Dai, P.; Bo, L.; Zhang, J.; and Zheng, Y. 2021. Traffic flow forecasting with spatial-temporal graph diffusion network. In Proceedings of the AAAI conference on artificial intelligence, volume 35, 15008--15015

    2021

  63. [71]

    Zhang, Y.; Gao, H.; Pei, J.; and Huang, H. 2022. Robust self-supervised structural graph neural network for social network prediction. In Proceedings of the ACM Web Conference 2022, 1352--1361

    2022

  64. [72]

    Zhou, K.; Liu, Z.; Qiao, Y.; Xiang, T.; and Loy, C. C. 2022. Domain generalization: A survey. IEEE Transactions on Pattern Analysis and Machine Intelligence

    2022

  65. [73]

    Zhuang, D.; Wang, S.; Koutsopoulos, H.; and Zhao, J. 2022. Uncertainty quantification of sparse travel demand prediction with spatial-temporal graph neural networks. In Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 4639--4647

    2022

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.