pith. sign in

arxiv: 2606.13571 · v1 · pith:5CJPE5SKnew · submitted 2026-06-11 · 💻 cs.LG · cs.AI

Existence Precedes Value: Joint Modeling of Observational Existence and Evolving States in Time Series Forecasting

Yifan Hu , Hongzhou Chen , Peiyuan Liu , Yiding Liu , Zewei Dong , Jiang-Ming Yang This is my paper

Pith reviewed 2026-06-27 07:22 UTC · model grok-4.3

classification 💻 cs.LG cs.AI
keywords time series forecastingmissing valuesobservabilityjoint modelingirregular samplingneural networks
0
0 comments X

The pith

Time series forecasting must jointly infer both whether future observations will exist and what their values will be.

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

Real-world time series data are frequently incomplete and irregular because of sensor dormancy, delays, or event-driven sampling. Prior approaches either impute missing values first or rely on continuous-time models, yet all of them presuppose that the timestamps of future valid observations are already known at inference time. Timeflies instead treats forecasting as a single joint task of predicting future observability and estimating the corresponding values. It maintains separate observation and value streams linked by three modules that embed reliability, guide dependencies, and produce joint predictions. On a new benchmark combining public and industrial data, the method shows consistent gains, demonstrating that explicit modeling of whether an observation will occur matters for accurate forecasts.

Core claim

Timeflies reformulates forecasting as the joint problem of future observability inference and value estimation by running an observation stream and a value stream that interact through reliability-aware embedding, observation-guided dependency modeling, and joint prediction.

What carries the argument

Dual observation and value streams coupled by reliability-aware embedding, observation-guided dependency modeling, and joint prediction modules.

Load-bearing premise

The three dedicated modules can capture the interaction between observation dynamics and state evolution without introducing new fitting artifacts.

What would settle it

A controlled comparison on held-out data with truly unknown future timestamps in which the joint model shows no accuracy gain over separate observability-then-value pipelines.

Figures

Figures reproduced from arXiv: 2606.13571 by Hongzhou Chen, Jiang-Ming Yang, Peiyuan Liu, Yiding Liu, Yifan Hu, Zewei Dong.

Figure 1
Figure 1. Figure 1: Evolution of Forecasting Paradigms for Time Series. (a) [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: The overall architecture of Timeflies. (1) Reliability-Aware Patch Embedding refines value tokens by incorporating patch-level reliability based on observed patterns and missingness in￾tervals. (2) Observation-Guided Value Attention integrates historical observation regularities into the attention mechanism to enhance value predictions. (3) Dual Prediction Head jointly predicts both future values and obser… view at source ↗
Figure 3
Figure 3. Figure 3: Visualization of Timeflies on a medium-irregular sample from ecom_BM. (a) Forecast￾ing trajectory: Filled and hollow circles denote valid and missing observations. Despite historical sparsity, the model closely follows the ground truth (green) with predictions (red). (b)–(e) Atten￾tion dynamics: Construction of the observation-conditioned routing mechanism, including (b) value attention, (c) observed atten… view at source ↗
Figure 4
Figure 4. Figure 4: (a) Sensitivity of Timeflies to the missingness loss weight η across datasets spanning four missing-ratio regimes. Parenthesized labels denote the missingness regime, and high (> 50%). (b) Model efficiency comparison under Bitbrains-FastStorage 1H dataset with high missing. • Focal Loss: Replacing the focal loss directly diminishes the existence inference capability (AUC), demonstrating its necessity in ha… view at source ↗
read the original abstract

Real-world time series are often highly incomplete and irregular due to sensor dormancy, transmission delays, and event-driven sampling, making reliable forecasting fundamentally challenging. Existing methods have evolved from impute-then-forecast pipelines to continuous-time models such as Neural ODEs and continuous-time graph networks. While these approaches improve the modeling of historical irregularity, they still rely on an implicit oracle assumption at inference time: the timestamps of future valid observations are presumed to be known in advance. This assumption limits practical relevance, since in many real systems the more fundamental question is not only what the future value will be, but also whether a valid observation will occur at all. In this paper, we propose Timeflies, a unified framework that reformulates forecasting as a joint problem of future observability inference and value estimation. To explicitly model the interaction between observation dynamics and state evolution, Timeflies adopts an observation stream and a value stream, coupled through three dedicated modules for reliability-aware embedding, observation-guided dependency modeling, and joint prediction. We further construct Shadow, a benchmark that combines natural missingness from public datasets with real-world industrial data, and introduce the Observation-Value Joint Entropy (OVJE) metric to comprehensively evaluate this coupled predictability. Extensive experiments show that Timeflies consistently outperforms existing methods, highlighting the importance of explicitly modeling future observability in time series forecasting with missing values. Code and dataset are available in https://github.com/ant-intl/Timeflies.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The paper claims that real-world time series forecasting is limited by an implicit oracle assumption of known future observation timestamps in existing methods (including Neural ODEs and continuous-time graph networks). It proposes Timeflies, which reformulates the task as joint future observability inference and value estimation via coupled observation and value streams with three modules (reliability-aware embedding, observation-guided dependency modeling, joint prediction). The work introduces the Shadow benchmark (combining public datasets with industrial data) and the OVJE metric, reporting consistent outperformance that highlights the value of explicitly modeling observational existence.

Significance. If the claimed gains are shown to arise specifically from the joint modeling rather than benchmark construction, metric choice, or unadjusted baselines, the result would be significant for practical deployment in domains with sensor dormancy and irregular sampling. The public release of code and the Shadow dataset strengthens reproducibility and enables follow-up work.

major comments (2)
  1. [Experiments] Experiments section: the central claim that Timeflies's advantage stems from the three dedicated modules requires explicit confirmation that baselines were adapted to predict observability (rather than retaining oracle access to future timestamps). Without this control, outperformance on Shadow/OVJE does not isolate the joint-modeling contribution.
  2. [Experiments] Experiments section: results should be reported under standard value-only metrics (MAE/RMSE) on the underlying public datasets in addition to OVJE. If gains vanish or become metric-specific under these controls, the necessity of the observation-value interaction modeling is not established.
minor comments (1)
  1. The abstract states that code and dataset are available, which is a positive contribution; the manuscript should include a direct link or repository reference in the main text as well.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback. The suggestions help clarify the experimental evidence for the joint modeling contribution. We address each major comment below.

read point-by-point responses

  1. Referee: [Experiments] Experiments section: the central claim that Timeflies's advantage stems from the three dedicated modules requires explicit confirmation that baselines were adapted to predict observability (rather than retaining oracle access to future timestamps). Without this control, outperformance on Shadow/OVJE does not isolate the joint-modeling contribution.

    Authors: We agree that isolating the joint-modeling contribution requires this control. The original baselines were run in their standard configurations, which retain oracle access to future timestamps as described in the manuscript's motivation. In the revision we will adapt the baselines to also predict observability (using the same OVJE metric and coupled prediction setup where feasible) and report the updated comparisons. This will directly test whether the gains arise from the three modules rather than task formulation. revision: yes

  2. Referee: [Experiments] Experiments section: results should be reported under standard value-only metrics (MAE/RMSE) on the underlying public datasets in addition to OVJE. If gains vanish or become metric-specific under these controls, the necessity of the observation-value interaction modeling is not established.

    Authors: We concur that standard value-only metrics provide an important additional control. The revised manuscript will include MAE and RMSE results on the public-dataset portions of Shadow (separate from the industrial data) for Timeflies and all baselines. These results will be presented alongside the OVJE numbers to assess whether advantages persist under conventional value forecasting evaluation. revision: yes

Circularity Check

0 steps flagged

No significant circularity detected in derivation or claims

full rationale

The paper describes Timeflies as a framework with three explicit modules (reliability-aware embedding, observation-guided dependency modeling, joint prediction) for joint observability and value forecasting, plus a new benchmark Shadow and OVJE metric. No equations, fitted parameters renamed as predictions, self-definitional steps, or load-bearing self-citations appear in the provided text. The central claim of outperformance rests on empirical experiments rather than reducing by construction to inputs or prior author work. This is a standard non-circular empirical modeling paper.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

No free parameters, axioms, or invented entities are specified in the abstract; the framework is described at the level of high-level modules without mathematical detail.

pith-pipeline@v0.9.1-grok · 5807 in / 1041 out tokens · 23926 ms · 2026-06-27T07:22:50.117965+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

48 extracted references · 5 canonical work pages · 2 internal anchors

  1. [1]

    Saits: Self-attention-based imputation for time series

    Wenjie Du, David Côté, and Yan Liu. Saits: Self-attention-based imputation for time series. Expert Systems with Applications, 219:119619, 2023

    2023

  2. [2]

    Revisiting multivariate time series forecasting with missing values.arXiv preprint arXiv:2509.23494, 2025

    Jie Yang, Yifan Hu, Kexin Zhang, Luyang Niu, Philip S Yu, and Kaize Ding. Revisiting multivariate time series forecasting with missing values.arXiv preprint arXiv:2509.23494, 2025

    work page arXiv 2025

  3. [3]

    Imputeformer: Low rankness- induced transformers for generalizable spatiotemporal imputation

    Tong Nie, Guoyang Qin, Wei Ma, Yuewen Mei, and Jian Sun. Imputeformer: Low rankness- induced transformers for generalizable spatiotemporal imputation. InProceedings of the 30th ACM SIGKDD conference on knowledge discovery and data mining, pages 2260–2271, 2024

    2024

  4. [4]

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

    Haoyi Zhou, Shanghang Zhang, Jieqi Peng, Shuai Zhang, Jianxin Li, Hui Xiong, and Wancai Zhang. Informer: Beyond efficient transformer for long sequence time-series forecasting. In Proceedings of the AAAI conference on artificial intelligence, volume 35, pages 11106–11115, 2021

    2021

  5. [5]

    Pe- riodicity decoupling framework for long-term series forecasting

    Tao Dai, Beiliang Wu, Peiyuan Liu, Naiqi Li, Jigang Bao, Yong Jiang, and Shu-Tao Xia. Pe- riodicity decoupling framework for long-term series forecasting. InInternational Conference on Learning Representations, 2024

    2024

  6. [6]

    Irregular multivariate time series forecasting: A transformable patching graph neural networks approach

    Weijia Zhang, Chenlong Yin, Hao Liu, Xiaofang Zhou, and Hui Xiong. Irregular multivariate time series forecasting: A transformable patching graph neural networks approach. InThe International Conference on Machine Learning, 2024

    2024

  7. [7]

    Hi-patch: Hierarchical patch GNN for irregular multivariate time series

    Yicheng Luo, Bowen Zhang, Zhen Liu, and Qianli Ma. Hi-patch: Hierarchical patch GNN for irregular multivariate time series. InThe International Conference on Machine Learning, 2025

    2025

  8. [8]

    The value of bat- tery energy storage in the continuous intraday market: Forecast vs

    Timothée Hornek, Youngsub Lee, Sergio Potenciano Menci, and Ivan Pavi´c. The value of bat- tery energy storage in the continuous intraday market: Forecast vs. perfect foresight strategies. In2025 IEEE Kiel PowerTech, pages 1–7. IEEE, 2025

    2025

  9. [9]

    Rethinking irregular time series forecasting: A simple yet effective baseline

    Xvyuan Liu, Xiangfei Qiu, Xingjian Wu, Zhengyu Li, Chenjuan Guo, Jilin Hu, and Bin Yang. Rethinking irregular time series forecasting: A simple yet effective baseline. InProceedings of the AAAI Conference on Artificial Intelligence, volume 40, pages 23873–23881, 2026

    2026

  10. [10]

    Transductive lstm for time-series prediction: An ap- plication to weather forecasting.Neural Networks, 125:1–9, 2020

    Zahra Karevan and Johan AK Suykens. Transductive lstm for time-series prediction: An ap- plication to weather forecasting.Neural Networks, 125:1–9, 2020

    2020

  11. [11]

    FinTSB: A Comprehensive and Practical Benchmark for Financial Time Series Forecasting

    Yifan Hu, Yuante Li, Peiyuan Liu, Yuxia Zhu, Naiqi Li, Tao Dai, Shu-tao Xia, Dawei Cheng, and Changjun Jiang. Fintsb: A comprehensive and practical benchmark for financial time series forecasting.arXiv preprint arXiv:2502.18834, 2025

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  12. [12]

    A short-term traffic flow prediction model based on an improved gate recurrent unit neural network.IEEE Transactions on Intelligent Transportation Systems, 23(9):16654–16665, 2021

    Wanneng Shu, Ken Cai, and Neal Naixue Xiong. A short-term traffic flow prediction model based on an improved gate recurrent unit neural network.IEEE Transactions on Intelligent Transportation Systems, 23(9):16654–16665, 2021

    2021

  13. [13]

    Heterogeneous multivariate time series imputation by transformer model with missing position encoding.Expert Systems with Applications, 271:126435, 2025

    Caizheng Liu, Zhengyu Zhu, Wanming Hao, and Gangcan Sun. Heterogeneous multivariate time series imputation by transformer model with missing position encoding.Expert Systems with Applications, 271:126435, 2025

    2025

  14. [14]

    Ginar+: A robust end-to-end framework for multivariate time series forecasting with missing values.IEEE Transactions on Knowledge and Data Engineering, 2025

    Chengqing Yu, Fei Wang, Zezhi Shao, Tangwen Qian, Zhao Zhang, Wei Wei, Zhulin An, Qi Wang, and Yongjun Xu. Ginar+: A robust end-to-end framework for multivariate time series forecasting with missing values.IEEE Transactions on Knowledge and Data Engineering, 2025

    2025

  15. [15]

    Analysis of missingness scenarios for observational health data.Journal of Personalized Medicine, 14(5):514, 2024

    Alireza Zamanian, Henrik von Kleist, Octavia-Andreea Ciora, Marta Piperno, Gino Lancho, and Narges Ahmidi. Analysis of missingness scenarios for observational health data.Journal of Personalized Medicine, 14(5):514, 2024

    2024

  16. [16]

    Data-driven causal analysis of observational biological time series.Elife, 11:e72518, 2022

    Alex Eric Yuan and Wenying Shou. Data-driven causal analysis of observational biological time series.Elife, 11:e72518, 2022. 10

    2022

  17. [17]

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

    Shiyang Li, Xiaoyong Jin, Yao Xuan, Xiyou Zhou, Wenhu Chen, Yu-Xiang Wang, and Xifeng Yan. Enhancing the locality and breaking the memory bottleneck of transformer on time series forecasting. InAdvances in Neural Information Processing Systems, volume 32, 2019

    2019

  18. [18]

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

    Haixu Wu, Jiehui Xu, Jianmin Wang, and Mingsheng Long. Autoformer: Decomposition transformers with auto-correlation for long-term series forecasting. InAdvances in Neural Information Processing Systems, volume 34, pages 22419–22430, 2021

    2021

  19. [19]

    Deep learning for time series forecast- ing: a survey.International Journal of Machine Learning and Cybernetics, 16(7):5079–5112, 2025

    Xiangjie Kong, Zhenghao Chen, Weiyao Liu, Kaili Ning, Lechao Zhang, Syauqie Muham- mad Marier, Yichen Liu, Yuhao Chen, and Feng Xia. Deep learning for time series forecast- ing: a survey.International Journal of Machine Learning and Cybernetics, 16(7):5079–5112, 2025

    2025

  20. [20]

    Seg- RNN: Segment recurrent neural network for long-term time series forecasting.IEEE Internet of Things Journal, 2025

    Shengsheng Lin, Weiwei Lin, Wentai Wu, Feiyu Zhao, Ruichao Mo, and Haotong Zhang. Seg- RNN: Segment recurrent neural network for long-term time series forecasting.IEEE Internet of Things Journal, 2025

    2025

  21. [21]

    Neural differential recurrent neural network with adaptive time steps.arXiv preprint arXiv:2306.01674, 2023

    Yixuan Tan, Liyan Xie, and Xiuyuan Cheng. Neural differential recurrent neural network with adaptive time steps.arXiv preprint arXiv:2306.01674, 2023

    work page arXiv 2023

  22. [22]

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

    Haixu Wu, Tengge Hu, Yong Liu, Hang Zhou, Jianmin Wang, and Mingsheng Long. TimesNet: Temporal 2d-variation modeling for general time series analysis. InInternational Conference on Learning Representations, 2023

    2023

  23. [23]

    MICN: Multi-scale local and global context modeling for long-term series forecasting

    Huiqiang Wang, Jian Peng, Feihu Huang, Jince Wang, Junhui Chen, and Yifei Xiao. MICN: Multi-scale local and global context modeling for long-term series forecasting. InInternational Conference on Learning Representations, 2022

    2022

  24. [24]

    FACT: Fine-grained across-variable convolution for multivariate time series forecasting

    Huiqiang Wang, Jieming Shi, and Li Qing. FACT: Fine-grained across-variable convolution for multivariate time series forecasting. InThe International Conference on Learning Repre- sentations, 2026

    2026

  25. [25]

    Are transformers effective for time series forecasting? InProceedings of the AAAI conference on artificial intelligence, volume 37, pages 11121–11128, 2023

    Ailing Zeng, Muxi Chen, Lei Zhang, and Qiang Xu. Are transformers effective for time series forecasting? InProceedings of the AAAI conference on artificial intelligence, volume 37, pages 11121–11128, 2023

    2023

  26. [26]

    TimeMixer: Decomposable multiscale mixing for time series forecasting

    Shiyu Wang, Haixu Wu, Xiaoming Shi, Tengge Hu, Huakun Luo, Lintao Ma, James Y Zhang, and Jun Zhou. TimeMixer: Decomposable multiscale mixing for time series forecasting. In International Conference on Learning Representations, 2024

    2024

  27. [27]

    OLinear: A linear model for time series forecasting in orthogonally transformed domain

    Wenzhen Yue, Yong Liu, Hao Wang, Haoxuan Li, Xianghua Ying, Ruohao Guo, Bowei Xing, and Ji Shi. OLinear: A linear model for time series forecasting in orthogonally transformed domain. InNeural Information Processing Systems, 2026

    2026

  28. [28]

    Adaptive multi-scale decompo- sition framework for time series forecasting.Proceedings of the AAAI Conference on Artificial Intelligence, 39(16):17359–17367, 2025

    Yifan Hu, Peiyuan Liu, Peng Zhu, Dawei Cheng, and Tao Dai. Adaptive multi-scale decompo- sition framework for time series forecasting.Proceedings of the AAAI Conference on Artificial Intelligence, 39(16):17359–17367, 2025

    2025

  29. [29]

    Bridging past and future: Distribution-aware alignment for time series forecasting

    Yifan Hu, Jie Yang, Tian Zhou, Peiyuan Liu, Yujin Tang, Rong Jin, and Liang Sun. Bridging past and future: Distribution-aware alignment for time series forecasting. InThe Fourteenth International Conference on Learning Representations, 2026

    2026

  30. [30]

    TimeFilter: Patch-specific spatial-temporal graph filtration for time series forecasting

    Yifan Hu, Guibin Zhang, Peiyuan Liu, Disen Lan, Naiqi Li, Dawei Cheng, Tao Dai, Shu-Tao Xia, and Shirui Pan. TimeFilter: Patch-specific spatial-temporal graph filtration for time series forecasting. InInternational Conference on Machine Learning, 2025

    2025

  31. [31]

    Connecting the dots: Multivariate time series forecasting with graph neural networks

    Zonghan Wu, Shirui Pan, Guodong Long, Jing Jiang, Xiaojun Chang, and Chengqi Zhang. Connecting the dots: Multivariate time series forecasting with graph neural networks. InPro- ceedings of the 26th ACM SIGKDD international conference on knowledge discovery & data mining, pages 753–763, 2020

    2020

  32. [32]

    Msgnet: Learning multi-scale inter-series correlations for multivariate time series forecasting.Proceedings of the AAAI Conference on Artificial Intelligence, 2024

    Wanlin Cai, Yuxuan Liang, Xianggen Liu, Jianshuai Feng, and Yuankai Wu. Msgnet: Learning multi-scale inter-series correlations for multivariate time series forecasting.Proceedings of the AAAI Conference on Artificial Intelligence, 2024. 11

    2024

  33. [33]

    FEDformer: Fre- quency enhanced decomposed transformer for long-term series forecasting

    Tian Zhou, Ziqing Ma, Qingsong Wen, Xue Wang, Liang Sun, and Rong Jin. FEDformer: Fre- quency enhanced decomposed transformer for long-term series forecasting. InInternational Conference on Machine Learning, pages 27268–27286. PMLR, 2022

    2022

  34. [34]

    Time- Bridge: Non-stationarity matters for long-term time series forecasting

    Peiyuan Liu, Beiliang Wu, Yifan Hu, Naiqi Li, Tao Dai, Jigang Bao, and Shu-tao Xia. Time- Bridge: Non-stationarity matters for long-term time series forecasting. InInternational Con- ference on Machine Learning, 2024

    2024

  35. [35]

    Nguyen, Phanwadee Sinthong, and Jayant Kalagnanam

    Yuqi Nie, Nam H. Nguyen, Phanwadee Sinthong, and Jayant Kalagnanam. A time series is worth 64 words: Long-term forecasting with transformers. InInternational Conference on Learning Representations, 2023

    2023

  36. [36]

    iTransformer: Inverted transformers are effective for time series forecasting

    Yong Liu, Tengge Hu, Haoran Zhang, Haixu Wu, Shiyu Wang, Lintao Ma, and Mingsheng Long. iTransformer: Inverted transformers are effective for time series forecasting. InInter- national Conference on Learning Representations, 2024

    2024

  37. [37]

    A survey of forecasting methods for irregular time series

    Xuanying Li, Sha Xiang, and Cheng Dai. A survey of forecasting methods for irregular time series. InIEEE International Conferences on Internet of Things, pages 83–90, 2025

    2025

  38. [38]

    Gru-ode-bayes: Continuous modeling of sporadically-observed time series.Advances in neural information processing systems, 32, 2019

    Edward De Brouwer, Jaak Simm, Adam Arany, and Yves Moreau. Gru-ode-bayes: Continuous modeling of sporadically-observed time series.Advances in neural information processing systems, 32, 2019

    2019

  39. [39]

    Latent ordinary differential equa- tions for irregularly-sampled time series.Advances in neural information processing systems, 32, 2019

    Yulia Rubanova, Ricky TQ Chen, and David K Duvenaud. Latent ordinary differential equa- tions for irregularly-sampled time series.Advances in neural information processing systems, 32, 2019

    2019

  40. [40]

    Modeling irregular time series with continuous recurrent units

    Mona Schirmer, Mazin Eltayeb, Stefan Lessmann, and Maja Rudolph. Modeling irregular time series with continuous recurrent units. InInternational conference on machine learning, pages 19388–19405. PMLR, 2022

    2022

  41. [41]

    Physiome-ODE: A Benchmark for Irregularly Sampled Multivariate Time Series Forecasting Based on Biological ODEs

    Christian Klötergens, Vijaya Krishna Yalavarthi, Randolf Scholz, Maximilian Stubbemann, Stefan Born, and Lars Schmidt-Thieme. Physiome-ode: A benchmark for irregularly sampled multivariate time series forecasting based on biological odes.arXiv preprint arXiv:2502.07489, 2025

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  42. [42]

    Biased temporal convolution graph net- work for time series forecasting with missing values

    Xiaodan Chen, Xiucheng Li, Bo Liu, and Zhijun Li. Biased temporal convolution graph net- work for time series forecasting with missing values. InInternational Conference on Learning Representations, 2024

    2024

  43. [43]

    Grafiti: Graphs for fore- casting irregularly sampled time series

    Vijaya Krishna Yalavarthi, Kiran Madhusudhanan, Randolf Scholz, Nourhan Ahmed, Johannes Burchert, Shayan Jawed, Stefan Born, and Lars Schmidt-Thieme. Grafiti: Graphs for fore- casting irregularly sampled time series. InProceedings of the AAAI Conference on Artificial Intelligence, pages 16255–16263, 2024

    2024

  44. [44]

    Gift-EV AL: A benchmark for general time series forecasting model evaluation.arXiv preprint arXiv:2410.10393, 2024

    Taha Aksu, Gerald Woo, Juncheng Liu, Xu Liu, Chenghao Liu, Silvio Savarese, Caiming Xiong, and Doyen Sahoo. Gift-EV AL: A benchmark for general time series forecasting model evaluation.arXiv preprint arXiv:2410.10393, 2024

    work page arXiv 2024

  45. [45]

    Crossformer: Transformer utilizing cross-dimension depen- dency for multivariate time series forecasting

    Yunhao Zhang and Junchi Yan. Crossformer: Transformer utilizing cross-dimension depen- dency for multivariate time series forecasting. InInternational Conference on Learning Rep- resentations, 2023

    2023

  46. [46]

    WPMixer: Efficient multi-resolution mixing for long-term time series forecasting

    Md Mahmuddun Nabi Murad, Mehmet Aktukmak, and Yasin Yilmaz. WPMixer: Efficient multi-resolution mixing for long-term time series forecasting. InProceedings of the AAAI conference on artificial intelligence, volume 39, pages 19581–19588, 2025

    2025

  47. [47]

    Pytorch: An imperative style, high-performance deep learning library

    Adam Paszke, Sam Gross, Francisco Massa, Adam Lerer, James Bradbury, Gregory Chanan, Trevor Killeen, Zeming Lin, Natalia Gimelshein, Luca Antiga, et al. Pytorch: An imperative style, high-performance deep learning library. InAdvances in Neural Information Processing Systems, volume 32, 2019

    2019

  48. [48]

    Freq.” denotes sampling frequency, “#Var

    Diederik P Kingma. Adam: A method for stochastic optimization. InInternational Conference on Learning Representations, 2015. 12 A Additional Experimental Details A.1 Dataset Details Tab. 4 provides a detailed characterization of theShadowbenchmark, which integrates 15 public datasets fromGIFT-Evalwith 16 proprietary e-commerce datasets spanning domains su...

    2015