Pith. sign in

REVIEW 4 major objections 6 minor 9 cited by

Time-VLM: Exploring Multimodal Vision-Language Models for Augmented Time Series Forecasting

T0 review · 4 major / 6 minor · reviewed 2026-08-09 · deepseek-v4-flash

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

Pith's one-line read Frozen vision-language embeddings improve few-shot time series forecasts.

desk verdict Time-VLM is a plausible engineering contribution whose abstract overclaims: the paper's own tables show the method losing on ECL and Traffic and on 5 of 8 zero-shot transfer tasks, but the architecture and ablations still deserve a serious referee. read the letter →

arxiv 2502.04395 v2 pith:E7BL2RS6 submitted 2025-02-06 cs.CV cs.LG

Siru Zhong , Weilin Ruan , Ming Jin , Huan Li , Qingsong Wen , Yuxuan Liang This is my paper · ORCID
classification cs.CVcs.LG
keywords timeseriesforecastingvision-languagemodelsmultimodallearningfew-shotzero-shotretrieval-augmentedimagingtextaugmentation
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

The paper proposes Time-VLM, a framework that forecasts time series by combining three views of the data: the raw temporal signal, an image made from it, and a short text description of its statistics. The central claim is that feeding these three views through a frozen vision-language model and fusing the resulting embeddings with temporal features yields more accurate forecasts than text-only or vision-only augmentation, with the largest gains when training data is scarce. If correct, this would mean that pretrained cross-modal knowledge acquired on natural images and captions can be reused for time series without fine-tuning the multimodal backbone, an efficiency win over methods that fine-tune large language models. The paper supports the claim with experiments on long-term and short-term benchmarks, few-shot and zero-shot transfers, and a 143M-parameter model that outperforms a 3.4B-parameter text-augmented baseline on most settings.

What carries the argument

The load-bearing mechanism is the frozen ViLT backbone, a lightweight vision-language transformer pretrained on COCO, combined with three modality transformers. The retrieval-augmented learner produces temporal patch embeddings refined by a memory bank with top-k cosine retrieval and a global multi-head self-attention summary. The vision-augmented learner converts the series into a 64 by 64 RGB image whose channels encode raw values, FFT magnitudes, and sine/cosine periodicity, then interpolates and normalizes the pixels to [0,255] for the VLM. The text-augmented learner concatenates task, domain, statistic, and image-description tokens. Cross-modal multi-head attention uses temporal features as queries and VLM embeddings as keys and values, followed by a gated sum; only these lightweight components and the prediction head are trainable while the VLM stays frozen.

What would settle it

Train the exact Time-VLM pipeline on the few-shot ETT benchmarks with the frozen VLM replaced by the same architecture with randomly initialized weights that are never updated. If the few-shot gains over a temporal-only baseline do not disappear, then the pretrained VLM's cross-modal knowledge is not what drives the improvement; the paper's own ViT plus BERT custom comparison falls short of this test because that baseline uses separately pretrained encoders, not destroyed joint alignment.

Watch

Extended reading notes

Core claim

Time-VLM's discovery, as the authors state it, is that a frozen pretrained vision-language model can act as a bridge that aligns temporal, visual, and textual representations of a time series, and that this alignment transfers to forecasting. Three learners feed the VLM: a retrieval-augmented learner that patches the series and maintains a memory bank of historical patch embeddings; a vision-augmented learner that converts the series into a multi-channel image via FFT, periodicity encoding, and multi-scale convolution; and a text-augmented learner that writes prompts containing statistics, periodicity, and domain context. The VLM's embeddings are fused with the temporal features through cross-modal attention and a gated mechanism, and a lightweight predictor outputs the forecast. The paper's headline evidence is in the few-shot (5% and 10% training data) and zero-shot cross-domain settings, where Time-VLM beats Time-LLM and other text-augmented baselines while using about one twentieth of the parameters.

Load-bearing premise

The framework assumes that a vision-language model pretrained on natural photographs and captions already knows something useful about synthetic images of time series and their text statistics, and that this knowledge transfers to forecasting, even though the paper's own visualization shows its pretrained image and text clusters sitting far from time series embeddings.

Editorial extensions

If this is right

  • Time-VLM reports lower MSE and MAE than Time-LLM on most few-shot (5% and 10%) and zero-shot ETT transfers while using roughly one twentieth of the parameters.
  • On the M4 benchmark, Time-VLM achieves the best SMAPE, MASE, and OWA among the compared methods, including N-HiTS and N-BEATS.
  • The framework operates on raw series alone: text prompts and images are generated from the input itself, so no external auxiliary data is required at inference time.
  • Long-term forecasting on Weather, ECL, and Traffic shows competitive but not uniformly best results, with the paper stating that full-shot performance slightly lags specialized unimodal models on ECL and Traffic.

Reading between the lines

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

  • The ablation (vision 9.0%, text 2.1%, temporal 35.6%) suggests that most of the few-shot gain may come from the retrieval-augmented temporal learner rather than from the VLM's cross-modal knowledge, a hypothesis the paper does not directly test with a VLM-free version of the full pipeline.
  • Because the paper's own UMAP shows COCO image and text embeddings isolated from time series embeddings, a natural extension is to pretrain the VLM on time series-specific image-text pairs, such as rendered series with generated captions, which the authors list as future work and which would directly test the transfer premise.
  • A testable extension is to vary the image rendering pipeline (line graph, recurrence plot, or the proposed FFT plus periodicity image) under a fixed VLM to see whether visual representation quality, rather than the VLM itself, drives the vision branch's 9.0% contribution.
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 proposes Time-VLM, a multimodal framework for time series forecasting that combines a retrieval-augmented temporal branch (RAL), a vision branch (VAL) that converts time series into images via FFT/periodicity encodings and multi-scale convolutions, and a text branch (TAL) that generates statistical and contextual descriptions. The three branches are processed by a frozen VLM (ViLT by default) and fused through cross-modal attention and gating, then trained end-to-end with MSE. The model is evaluated on ETT, Weather, ECL, Traffic, and M4 benchmarks under full-shot, few-shot (5%/10%), and zero-shot (cross-domain ETT) settings. The paper claims superior performance, especially in few-shot and zero-shot scenarios, and attributes the gains to pre-trained VLM knowledge.

Significance. The idea of exploiting a frozen VLM to unify temporal, visual, and textual modalities is timely and original, and the released code plus the custom ViT+BERT control (Table 8) are good scientific practices. If the empirical claims held, the framework would be a useful addition to the multimodal forecasting toolbox, with the practical advantage of being much smaller than Time-LLM. However, the current evidence only partially supports the headline claims: the gains are concentrated on the ETT and Weather datasets, while Time-VLM loses to Time-LLM on ECL and Traffic in both few-shot and full-shot settings, and the zero-shot superiority is not consistently observed. The paper's own ablations indicate that the temporal RAL module, not the VLM-based vision/text branches, is the main contributor to accuracy.

major comments (4)
  1. [Section 4.2, Table 3] The text states that Time-VLM 'consistently outperforms or matching' baselines in zero-shot, but against Time-LLM it wins on only 3 of 8 ETT transfer tasks (ETTh1→ETTh2, ETTm1→ETTh2, ETTm1→ETTm2) and loses on the remaining 5, sometimes by large margins (e.g., ETTh1→ETTm2 MSE 0.293 vs 0.273; ETTh2→ETTh1 MSE 0.496 vs 0.479). The abstract's claim of 'superior performance, particularly in zero-shot' is therefore contradicted by the full results in Table 14.
  2. [Section 4.1, Tables 1-2] In few-shot settings, Time-VLM is best on the ETT datasets and Weather but is worse than Time-LLM (and several other baselines) on ECL and Traffic; e.g., at 5% training, ECL MSE is 0.218 vs 0.179 for Time-LLM and Traffic MSE is 0.558 vs 0.423 for Time-LLM. The paper's summary that Time-VLM 'consistently outperforms most baselines across datasets' is technically true only if 'most' is read as 'more than half,' but the claimed 'particularly in few-shot' superiority is not supported on two of the seven datasets.
  3. [Section 4, Appendices A-B] No error bars, confidence intervals, or significance tests are reported for any experiment. Many reported improvements are small (e.g., Weather full-shot MSE 0.224 vs 0.225 for Time-LLM; M4 overall SMAPE 11.894 vs 11.983), so without variance estimates or multiple seeds, the reader cannot determine whether the observed differences are real or within run-to-run noise.
  4. [Section 4.5, Table 6 and Table 8] The ablations attribute 35.6% MSE degradation to removing RAL, 9.0% to removing VAL, and 2.1% to removing TAL. Additionally, replacing the VLM with separate ViT+BERT encoders costs only 0.012 MSE (0.348 vs 0.336) on ETTh2. These results weaken the paper's central claim that the VLM's pre-trained cross-modal alignment is the key driver of performance; the evidence suggests the retrieval-augmented temporal branch is the dominant contributor. The discussion should be reweighted accordingly.
minor comments (6)
  1. [Abstract and Section 5] The limitations acknowledged in the paper (full-shot performance lags on ECL/Traffic; TAL has limited impact; irregular-pattern datasets are harder) are not reflected in the abstract or conclusion, which claim 'superior performance' and 'a new direction'; the claims should be toned down to match the evidence.
  2. [Figure 4 caption] The caption contains a duplicated phrase 'of of' ('of of multimodal and temporal embeddings').
  3. [Appendix B.2, Table 14] The ETTm2→ETTh2 zero-shot transfer appears to miss the 192-horizon row, making the average computed over three horizons rather than four.
  4. [Equation (2)] The notation topk(E^{(i)}_p) is used before E^{(i)}_p is defined; consider defining the per-sample patch embedding explicitly.
  5. [Section 3.2] The sentence introducing bilinear interpolation is followed by equation (8), which uses I(x_i, y_j) for the four nearest neighbors; clarifying this notation would improve readability.
  6. [Section 4, Baselines] The reference to '(2024a)' in the baseline description is incomplete; please cite the actual source.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: Time-VLM is an end-to-end trained forecasting model whose central premise is tested by ablations and external benchmarks; the few cited works by co-authors are baselines or background, not load-bearing derivations.

full rationale

The paper does not derive its forecasting claim from a fitted parameter that is then called a prediction. Section 3.5 defines the objective as direct MSE on forecast error (Eq. 15), and all components (RAL, VAL, TAL) transform the same input time series; none encodes the target or a fitted summary of the reported wins. The central premise that a frozen VLM's cross-modal alignment helps is empirically tested rather than assumed: Table 6 shows removing VAL costs 9.0% MSE and removing TAL costs 2.1%, and Table 8 shows a custom ViT+BERT combination (0.348 MSE) underperforms the pretrained ViLT (0.336 MSE) on ETTh2. Comparisons to Time-LLM, GPT4TS, PatchTST, etc., are external benchmarks under a shared protocol, so the reported results are not forced by construction. The paper contains several self-citations (Time-LLM, UniTime, Time-FFM, and the Liang et al. survey), but they are used as baselines, related work, or background context, not as the sole justification for the central claim; there is no uniqueness theorem or fitted input imported from those works. The abstract's 'superior performance, particularly in few-shot and zero-shot' is questionably supported by the full tables (e.g., in Table 3 Time-VLM wins only 3 of 8 ETT transfer tasks on MSE against Time-LLM, and in Tables 1-2 it loses on ECL and Traffic), but that is an overclaim or empirical-support issue, not circularity. Accordingly the circularity score is 0.

Assumptions & free parameters 6 free parameters · 4 assumptions · 0 invented entities

The central empirical results depend on a set of tuned hyperparameters (periodicity, normalization constant, sequence length, dimensions) and on domain assumptions about periodic structure and VLM transfer. None of these are derived or externally benchmarked, so the measurement of 'multimodal benefit' is sensitive to them.

free parameters (6)
  • Periodicity P (per dataset) = 24/96/144, and M4 values 1,3,4,24
    Used in sine/cosine periodicity encoding (Eq. 7); values are set per dataset from known sampling frequency, and performance depends on this choice (Appendix A.1.2).
  • Normalization constant = 0.4
    Applied to stabilize training; hyperparameter study shows peak at 0.4 (Figure 5).
  • Input sequence length = 512
    Selected from hyperparameter study; performance varies with length, 512 optimal for most datasets.
  • Model dimension d_model = 128
    Default architecture parameter; study shows dataset-dependent optimal values 128-512.
  • Gate fusion dimension = 256
    Tuned per task; 256 for medium-range, 336/512 for harder settings.
  • Image size = 64x64
    Chosen to balance efficiency and information preservation, not derived from data.
assumptions (4)
  • domain assumption FFT and sine/cosine periodic encodings expose forecasting-relevant frequency and phase information in these datasets
    Used in Section 3.2, Eqs. (6)-(7); standard in signal processing and prior time series imaging, but the specific benefit is not isolated here.
  • domain assumption Dataset-specific periodicity P is known in advance and reflects the dominant cycle (24, 96, 144)
    Appendix A.1.2 assigns P from sampling frequency; a wrong P would inject a misleading periodic bias into the generated image.
  • domain assumption Pretrained VLM cross-modal alignment transfers to synthetic time-series images and generated statistics
    Core motivation stated in Sections 1 and 4.5; only weakly supported by small ablation effects (w/o VAL +9%, w/o TAL +2.1%).
  • domain assumption Min-max normalization of generated images to [0,255] preserves forecasting-relevant magnitude information
    Eq. (9) in Section 3.2; scaling that discards relative magnitudes may remove information needed for forecast levels.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Time-VLM: Exploring Multimodal Vision-Language Models for Augmented Time Series Forecasting." pith.science (2026). https://pith.science/paper/E7BL2RS6

@misc{pith2026250204395,
  author       = {Pith},
  title        = {Pith review of: Time-VLM: Exploring Multimodal Vision-Language Models for Augmented Time Series Forecasting},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/E7BL2RS6}},
  note         = {Machine review of arXiv:2502.04395}
}
read the original abstract

Recent advancements in time series forecasting have explored augmenting models with text or vision modalities to improve accuracy. While text provides contextual understanding, it often lacks fine-grained temporal details. Conversely, vision captures intricate temporal patterns but lacks semantic context, limiting the complementary potential of these modalities. To address this, we propose \method, a novel multimodal framework that leverages pre-trained Vision-Language Models (VLMs) to bridge temporal, visual, and textual modalities for enhanced forecasting. Our framework comprises three key components: (1) a Retrieval-Augmented Learner, which extracts enriched temporal features through memory bank interactions; (2) a Vision-Augmented Learner, which encodes time series as informative images; and (3) a Text-Augmented Learner, which generates contextual textual descriptions. These components collaborate with frozen pre-trained VLMs to produce multimodal embeddings, which are then fused with temporal features for final prediction. Extensive experiments demonstrate that Time-VLM achieves superior performance, particularly in few-shot and zero-shot scenarios, thereby establishing a new direction for multimodal time series forecasting. Code is available at https://github.com/CityMind-Lab/ICML25-TimeVLM.

Figures

Figures reproduced from arXiv: 2502.04395 by the authors.

Figure 1
Figure 1. Our Time-VLM combines text (Right) and vision (Left) modalities to augment time series forecasting. (Idrees et al., 2019), climate (Karevan & Suykens, 2020), en￾ergy (Deb et al., 2017), and transportation (Zheng & Huang, 2020). Accurate forecasting supports proactive risk mitiga￾tion, efficient resource allocation, and data-driven decision￾making. Traditional models like ARIMA, while historically dominant, struggle … view at source ↗
Figure 2
Figure 2. Overview of the Time-VLM framework. 3. Methodology To address the limitations of single-modality approaches and leverage the complementary strengths of visual, tex￾tual, and temporal modalities, we propose Time-VLM, a unified framework that integrates these modalities for en￾hanced time series forecasting. As illustrated in [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 4
Figure 4. Interpretability visualization of Time-VLM: mul￾timodal feature alignment via UMAP. pure image representations retain modality-specific charac￾teristics. Similarly, COCO-Text forms a completely separate cluster, highlighting significant modality gaps. In contrast, COCO-Pair exhibits maximal overlap with time-series data, demonstrating strong cross-modal complementarity. The textual semantics in COCO-Pair bridge visu… view at source ↗
Figures from the paper (4 more)
Figure 3
Figure 3. Figure 3: UMAP visualization (left) and gate weight distri [PITH_FULL_IMAGE:figures/full_fig_p008_3.png]
Figure 5
Figure 5. Figure 5: Hyperparameters sensitivity analysis on input [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]
Figure 6
Figure 6. Figure 6: Time series transformed images, capturing key temporal characteristics, including trends, stationarity, seasonality, [PITH_FULL_IMAGE:figures/full_fig_p017_6.png]
Figure 10
Figure 10. Figure 10: Prediction results visualization for ETTh1, ETTm1, ECL, and Traffic datasets at 720 prediction lengths. True [PITH_FULL_IMAGE:figures/full_fig_p019_10.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 9 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Overcoming the Modality Gap in Context-Aided Forecasting

    cs.LG 2026-03 unverdicted novelty 7.0 of 10

    A semi-synthetic dataset of 7 million context-augmented time series windows with verifier-filtered contexts enables transfer to real-world context-aided forecasting and suggests data quality—not architecture—was the b...

  2. Watermarking Large Language Model-based Time Series Forecasting

    cs.IR 2025-07 conditional novelty 7.0 of 10

    Waltz embeds watermarks into LLM-based time series forecasts by nudging a few patch embeddings toward 'cold' LLM tokens, and detects them with a z-score test.

  3. TSRouter: Dynamic Modality-Model Selection for Time Series Reasoning

    cs.LG 2026-07 conditional novelty 6.5 of 10

    A heterogeneous-graph router jointly selects the optimal modality (text, vision, or both) and model per time series query, beating prior routing baselines and generalizing to unseen models and tasks.

  4. From Images to Signals: Are Large Vision Models Useful for Time Series Analysis?

    cs.LG 2025-05 conditional novelty 6.0 of 10

    Large vision models slightly beat strong baselines on imaged time series classification, but their forecasting advantage is narrow, tied to periodic patterns, and shrinks with long histories.

  5. Teaching Time Series to See and Speak: Forecasting with Aligned Visual and Textual Perspectives

    cs.LG 2025-06 reject novelty 5.0 of 10

    TimesCLIP aligns image-based and text-based views of the same time series via contrastive learning to improve forecasting accuracy on several benchmarks, but the full multimodal model is not used on two of the six lon...

  6. XFMNet: Decoding Cross-Site and Nonstationary Water Patterns via Stepwise Multimodal Fusion for Long-Term Water Quality Forecasting

    cs.LG 2025-08 unverdicted novelty 4.0 of 10

    XFMNet fuses local water-quality time series with remote-sensing precipitation imagery through stepwise multimodal fusion to improve long-term, multi-site water quality forecasting.

  7. Diffusion Models for Time Series Forecasting: A Survey

    stat.ML 2025-07 conditional novelty 4.0 of 10

    A survey classifies diffusion-based time series forecasting models into a two-axis taxonomy by conditioning source and integration method.

  8. Large Language Model Powered Intelligent Urban Agents: Concepts, Capabilities, and Applications

    cs.MA 2025-07 conditional novelty 4.0 of 10

    The paper defines urban LLM agents, surveys their sensing, memory, reasoning, execution, and learning workflows, and organizes their applications across planning, transportation, environment, safety, and society.

  9. CLIPTime: Time-Aware Multimodal Representation Learning from Images and Text

    cs.CV 2025-08 reject novelty 2.0 of 10

    CLIPTime adds a classification head and a transformer-style regression head to CLIP embeddings, hitting 98.7% accuracy on synthetic fungi but with weak timestamp predictions, especially for spores.

Reference graph

Works this paper leans on

39 extracted references · 23 canonical work pages · cited by 9 Pith papers

  1. [1]

    G., Oreshkin, B

    Challu, C., Olivares, K. G., Oreshkin, B. N., Ramirez, F. G., Canseco, M. M., and Dubrawski, A. Nhits: neural hierarchical interpolation for time series forecasting. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 37, pp.\ 6989--6997, 2023

    work page 2023

  2. [2]

    Llm4ts: Two-stage fine-tuning for time-series forecasting with pre-trained llms

    Chang, C., Peng, W.-C., and Chen, T.-F. Llm4ts: Two-stage fine-tuning for time-series forecasting with pre-trained llms. arXiv preprint arXiv:2308.08469, 2023

    arXiv 2023

  3. [3]

    J., Sun, J., and Liu, C

    Chen, M., Shen, L., Li, Z., Wang, X. J., Sun, J., and Liu, C. Visionts: Visual masked autoencoders are free-lunch zero-shot time series forecasters, 2024. URL https://arxiv.org/abs/2408.17253

    arXiv 2024

  4. [4]

    E., and Shah, K

    Deb, C., Zhang, F., Yang, J., Lee, S. E., and Shah, K. W. A review on time series forecasting techniques for building energy consumption. Renewable and Sustainable Energy Reviews, 74: 0 902--924, 2017

    work page 2017

  5. [5]

    Gruver, N., Finzi, M., Qiu, S., and Wilson, A. G. Large language models are zero-shot time series forecasters. In Advances in Neural Information Processing Systems, 2023

    work page 2023

  6. [6]

    M., Alam, M

    Idrees, S. M., Alam, M. A., and Agarwal, P. A prediction approach for stock market volatility based on time series data. IEEE Access, 7: 0 17287--17298, 2019

    work page 2019

  7. [7]

    Scaling up visual and vision-language representation learning with noisy text supervision

    Jia, C., Yang, Y., Xia, Y., Chen, Y.-T., Parekh, Z., Pham, H., Le, Q., Sung, Y.-H., Li, Z., and Duerig, T. Scaling up visual and vision-language representation learning with noisy text supervision. In International conference on machine learning, pp.\ 4904--4916. PMLR, 2021

    2021

  8. [8]

    Y., Shi, X., Chen, P.-Y., Liang, Y., Li, Y.-F., Pan, S., and Wen, Q

    Jin, M., Wang, S., Ma, L., Chu, Z., Zhang, J. Y., Shi, X., Chen, P.-Y., Liang, Y., Li, Y.-F., Pan, S., and Wen, Q. Time-LLM : Time series forecasting by reprogramming large language models. In International Conference on Learning Representations (ICLR), 2024

    2024

Show all 39 references
  1. [9]

    and Suykens, J

    Karevan, Z. and Suykens, J. A. Transductive lstm for time-series prediction: An application to weather forecasting. Neural Networks, 125: 0 1--9, 2020

    2020

  2. [10]

    Vilt: Vision-and-language transformer without convolution or region supervision

    Kim, W., Son, B., and Kim, I. Vilt: Vision-and-language transformer without convolution or region supervision. In International conference on machine learning, pp.\ 5583--5594. PMLR, 2021

    2021

  3. [11]

    Reformer: The efficient transformer

    Kitaev, N., Kaiser, ., and Levskaya, A. Reformer: The efficient transformer. In International Conference on Learning Representations, 2020

    2020

  4. [12]

    Modeling long-and short-term temporal patterns with deep neural networks

    Lai, G., Chang, W.-C., Yang, Y., and Liu, H. Modeling long-and short-term temporal patterns with deep neural networks. In The 41st international ACM SIGIR conference on research & development in information retrieval, pp.\ 95--104, 2018

    2018

  5. [13]

    Blip-2: Bootstrapping language-image pre-training with frozen image encoders and large language models

    Li, J., Li, D., Savarese, S., et al. Blip-2: Bootstrapping language-image pre-training with frozen image encoders and large language models. In International conference on machine learning, pp.\ 19730--19742. PMLR, 2023

    2023

  6. [14]

    Enhancing the locality and breaking the memory bottleneck of transformer on time series forecasting

    Li, S., Jin, X., Xuan, Y., Zhou, X., Chen, W., Wang, Y.-X., and Yan, X. Enhancing the locality and breaking the memory bottleneck of transformer on time series forecasting. In Advances in neural information processing systems, 2019

    2019

  7. [15]

    Forecasting with time series imaging

    Li, X., Kang, Y., and Li, F. Forecasting with time series imaging. Expert Systems with Applications, 160: 0 113680, 2020

    2020

  8. [16]

    Foundation models for time series analysis: A tutorial and survey

    Liang, Y., Wen, H., Nie, Y., Jiang, Y., Jin, M., Song, D., Pan, S., and Wen, Q. Foundation models for time series analysis: A tutorial and survey. In Proceedings of the 30th ACM SIGKDD conference on knowledge discovery and data mining, pp.\ 6555--6565, 2024

    2024

  9. [17]

    Liu, H., Li, C., Wu, Q., and Lee, Y. J. Visual instruction tuning. arXiv preprint arXiv:2304.08485, 2023

    2023 arXiv

  10. [18]

    Time-ffm: Towards lm-empowered federated foundation model for time series forecasting

    Liu, Q., Liu, X., Liu, C., Wen, Q., and Liang, Y. Time-ffm: Towards lm-empowered federated foundation model for time series forecasting. arXiv preprint arXiv:2405.14252, 2024 a

    2024 arXiv

  11. [19]

    X., and Dustdar, S

    Liu, S., Yu, H., Liao, C., Li, J., Lin, W., Liu, A. X., and Dustdar, S. Pyraformer: Low-complexity pyramidal attention for long-range time series modeling and forecasting. In International Conference on Learning Representations, 2022 a

    2022

  12. [20]

    Unitime: A language-empowered unified model for cross-domain time series forecasting

    Liu, X., Hu, J., Li, Y., Diao, S., Liang, Y., Hooi, B., and Zimmermann, R. Unitime: A language-empowered unified model for cross-domain time series forecasting. In Proceedings of the ACM on Web Conference 2024, pp.\ 4095--4106, 2024 b

    2024

  13. [21]

    Non-stationary transformers: Exploring the stationarity in time series forecasting

    Liu, Y., Wu, H., Wang, J., and Long, M. Non-stationary transformers: Exploring the stationarity in time series forecasting. Advances in Neural Information Processing Systems, 35: 0 9881--9893, 2022 b

    2022

  14. [22]

    The m4 competition: Results, findings, conclusion and way forward

    Makridakis, S., Spiliotis, E., and Assimakopoulos, V. The m4 competition: Results, findings, conclusion and way forward. International Journal of Forecasting, 34 0 (4): 0 802--808, 2018

    2018

  15. [23]

    R., Jain, L., et al

    Medsker, L. R., Jain, L., et al. Recurrent neural networks. Design and Applications, 5 0 (64-67): 0 2, 2001

    2001

  16. [24]

    H., Sinthong, P., and Kalagnanam, J

    Nie, Y., Nguyen, N. H., Sinthong, P., and Kalagnanam, J. A time series is worth 64 words: Long-term forecasting with transformers. In International Conference on Learning Representations, 2023

    2023

  17. [25]

    N., Carpov, D., Chapados, N., and Bengio, Y

    Oreshkin, B. N., Carpov, D., Chapados, N., and Bengio, Y. N-beats: Neural basis expansion analysis for interpretable time series forecasting. In International Conference on Learning Representations, 2020

    2020

  18. [26]

    W., Hallacy, C., Ramesh, A., Goh, G., Agarwal, S., Sastry, G., Askell, A., Mishkin, P., Clark, J., et al

    Radford, A., Kim, J. W., Hallacy, C., Ramesh, A., Goh, G., Agarwal, S., Sastry, G., Askell, A., Mishkin, P., Clark, J., et al. Learning transferable visual models from natural language supervision. In International conference on machine learning, pp.\ 8748--8763. PMLR, 2021

    2021

  19. [27]

    Visual time series forecasting: an image-driven approach

    Sood, S., Zeng, Z., Cohen, N., Balch, T., and Veloso, M. Visual time series forecasting: an image-driven approach. In Proceedings of the Second ACM International Conference on AI in Finance, pp.\ 1--9, 2021

    2021

  20. [28]

    Timemixer++: A general time series pattern machine for universal predictive analysis

    Wang, S., Li, J., Shi, X., Ye, Z., Mo, B., Lin, W., Ju, S., Chu, Z., and Jin, M. Timemixer++: A general time series pattern machine for universal predictive analysis. arXiv preprint arXiv:2410.16032, 2024

    2024 arXiv

  21. [29]

    Etsformer: Exponential smoothing transformers for time-series forecasting

    Woo, G., Liu, C., Sahoo, D., Kumar, A., and Hoi, S. Etsformer: Exponential smoothing transformers for time-series forecasting. In arXiv preprint arXiv:2202.01381, 2022

    2022 arXiv

  22. [30]

    Autoformer: Decomposition transformers with auto-correlation for long-term series forecasting

    Wu, H., Xu, J., Wang, J., and Long, M. Autoformer: Decomposition transformers with auto-correlation for long-term series forecasting. Advances in Neural Information Processing Systems, 34: 0 22419--22430, 2021

    2021

  23. [31]

    Timesnet: Temporal 2d-variation modeling for general time series analysis

    Wu, H., Hu, T., Liu, Y., Zhou, H., Wang, J., and Long, M. Timesnet: Temporal 2d-variation modeling for general time series analysis. In International Conference on Learning Representations, 2023 a

    2023

  24. [32]

    Timesnet: Temporal 2d-variation modeling for general time series analysis

    Wu, H., Hu, T., Liu, Y., Zhou, H., Wang, J., and Long, M. Timesnet: Temporal 2d-variation modeling for general time series analysis. In International Conference on Learning Representations, 2023 b

    2023

  25. [33]

    Are transformers effective for time series forecasting? In Proceedings of the AAAI conference on artificial intelligence, volume 37, pp.\ 11121--11128, 2023

    Zeng, A., Chen, M., Zhang, L., and Xu, Q. Are transformers effective for time series forecasting? In Proceedings of the AAAI conference on artificial intelligence, volume 37, pp.\ 11121--11128, 2023

    2023

  26. [34]

    Less is more: Fast multivariate time series forecasting with light sampling-oriented mlp structures

    Zhang, T., Zhang, Y., Cao, W., Bian, J., Yi, X., Zheng, S., and Li, J. Less is more: Fast multivariate time series forecasting with light sampling-oriented mlp structures. arXiv preprint arXiv:2207.01186, 2022

    2022 arXiv

  27. [35]

    and Huang, M

    Zheng, J. and Huang, M. Traffic flow forecast through time series analysis based on deep learning. IEEE Access, 8: 0 82562--82570, 2020

    2020

  28. [36]

    Informer: Beyond efficient transformer for long sequence time-series forecasting

    Zhou, H., Zhang, S., Peng, J., Zhang, S., Li, J., Xiong, H., and Zhang, W. Informer: Beyond efficient transformer for long sequence time-series forecasting. In Proceedings of the AAAI conference on Artificial Intelligence, pp.\ 11106--11115, 2021

    2021

  29. [37]

    Fedformer: Frequency enhanced decomposed transformer for long-term series forecasting

    Zhou, T., Ma, Z., Wen, Q., Wang, X., Sun, L., and Jin, R. Fedformer: Frequency enhanced decomposed transformer for long-term series forecasting. In International Conference on Machine Learning, pp.\ 27268--27286, 2022

    2022

  30. [38]

    One fits all: Power general time series analysis by pretrained lm

    Zhou, T., Niu, P., Wang, X., Sun, L., and Jin, R. One fits all: Power general time series analysis by pretrained lm. In Advances in Neural Information Processing Systems, 2023

    2023

  31. [39]

    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 gl...

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.