Pith. sign in

REVIEW 4 major objections 4 minor 44 references

TSDS-Toolbox: A Toolbox for Measuring Time-Series Dataset Similarity

T0 review · 4 major / 4 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read TSDS-Toolbox standardizes time-series dataset similarity benchmarking, and its experiments indicate that no single similarity metric consistently predicts downstream task performance.

desk verdict A genuinely useful integration toolbox whose headline empirical claim about similarity metrics is undermined by a pooling confound. read the letter →

arxiv 2608.08119 v1 pith:MAYN43DP submitted 2026-08-08 cs.LG

Yen-Ku Liu , Hongjie Chen , Ryan A. Rossi , Franck Dernoncourt This is my paper · ORCID
classification cs.LG
keywords time-seriesdatasetsimilaritybenchmarkingframeworkmetricsfoundationmodelssourceselectiontransferlearningforecastingdynamictimewarping
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

This paper presents TSDS-Toolbox, a configuration-driven software framework for measuring and benchmarking similarity between time-series datasets. It brings dataset-level metrics (including Wasserstein distance, maximum mean discrepancy, optimal transport, and match-and-deform) and reducer-based series-level metrics into one pipeline, then evaluates how well each metric predicts performance on downstream forecasting and classification tasks. The paper's central empirical claim is that no single similarity method consistently aligns with downstream task performance; the best metric depends on the task and on the forecasting model used. If this claim is right, the toolbox gives researchers a practical, reproducible way to choose source datasets and to compare similarity metrics for transfer learning instead of relying on a single default measure.

What carries the argument

The central mechanism is a four-layer pipeline. A Data Layer converts heterogeneous time-series inputs into a common representation with consistent preprocessing; a Similarity Layer computes pairwise distance matrices for both dataset-level metrics and series-level metrics, where series-level metrics first apply a reducer (DTW Barycenter Averaging or PCA) to map each dataset to a representative sequence and then apply a series distance (DTW, Euclidean distance, or longest common subsequence). An Evaluation Layer measures whether those distances carry information about downstream performance, using a one-class k-NN membership rate for classification and fine-tuned foundation-model forecast MSE for forecasting, in both out-of-distribution and transfer-learning settings. An Analysis Layer then computes Pearson correlations between the distance matrices and the task losses, along with heatmap and network-graph visualizations, which is the step that makes different similarity methods comparable on equal footing.

What would settle it

Repeat the comparison on a much larger and more diverse set of public time-series datasets, or measure utility by full end-task accuracy after fine-tuning instead of Pearson correlation with two proxy losses; if one similarity metric then ranks first across classification, out-of-distribution forecasting, and transfer-learning forecasting, the paper's central claim would be falsified.

Watch

Extended reading notes

Core claim

The central discovery is that time-series dataset similarity is task-dependent, and the toolbox is designed to make that dependency visible and measurable. Across the evaluated datasets and models, Pearson correlations between dataset-distance scores and downstream losses change substantially from one setting to another: dataset-level metrics such as match-and-deform and Wasserstein distance align most strongly with out-of-distribution forecasting loss when paired with one forecasting foundation model, while optimal transport aligns best with the one-class classification proxy, and no metric is consistently top-ranked across all columns. The experiments also show that DBA-based reduction generally tracks downstream performance better than PCA-based reduction, and that LCSS and MMD show weak alignment throughout. The paper concludes that dataset similarity should be assessed per task and that a standardized benchmarking environment is the appropriate basis for that assessment.

Load-bearing premise

The paper's empirical conclusion rests on assuming that the correlation between a similarity score and two chosen performance measures (a classification membership rate and forecast error) on 25 sampled datasets is a complete measure of a metric's usefulness; if that assumption fails, the claim that no metric is consistently better is unsupported.

Editorial extensions

If this is right

  • Researchers can now compare any newly proposed dataset-similarity metric against existing ones under identical data preprocessing, reducers, and downstream evaluation, making benchmark results directly comparable across papers.
  • In forecasting out-of-distribution settings, dataset-level metrics (match-and-deform and Wasserstein distance) are reasonable first choices for source-dataset selection, based on their higher correlations in the experiments.
  • DBA reducers are generally preferable to PCA reducers for downstream alignment in these experiments, suggesting that averaging-based reduction retains more task-relevant information than projection-based reduction.
  • LCSS and MMD show weak alignment with downstream performance, so users should be cautious about using them as proxies for transferability.
  • The task-dependence result implies that source-dataset selection pipelines should evaluate candidate similarity metrics on the actual downstream task rather than trusting one metric a priori.

Reading between the lines

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

  • A natural extension of the toolbox's correlation tables is to learn a meta-model that, given a target task and model, predicts which similarity metric will align best with downstream loss; the paper stops at reporting the rankings, but its modular design makes this straightforward.
  • Since the empirical conclusion is based on only 25 datasets and two forecasting models, the same protocol should be run on a larger and more heterogeneous collection; the observed task-dependence may turn out to be an artifact of the small sample, or it may sharpen into a more concrete rule about which metric fits which task.
  • A user could run the toolbox offline on their own domain's datasets to pick a similarity metric before doing transfer learning, effectively turning the paper's comparative results into a per-domain recommendation procedure that the paper does not itself propose.
  • One could extend the evaluation layer to regression, anomaly detection, or generation tasks; if the task-dependence result holds there, dataset similarity would be even more clearly a per-task design choice rather than a property of the data alone.
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 / 4 minor

Summary. The paper presents TSDS-Toolbox, a configuration-driven Python framework for computing pairwise time-series dataset similarity using dataset-level metrics (WSD, MMD, OT, MAD) and reducer-based series-level metrics (DBA/PCA reducers with DTW, ED, LCSS distances). It also provides downstream evaluation pipelines for classification and forecasting, and uses Pearson correlation between similarity distances and task losses to assess which metrics align with transfer performance. Experiments on 25 GluonTS datasets indicate that no single similarity metric consistently outperforms others across tasks, and that reducer choice matters. The code is publicly available.

Significance. If the toolbox is robust and the empirical findings hold, it would provide a valuable standardized benchmarking resource for time-series dataset similarity, addressing a real fragmentation in the area. The paper ships code and a configurable pipeline, which is a strength. The empirical claim about task-dependence of similarity metrics is potentially useful but is currently under-supported by the presented analysis.

major comments (4)
  1. [Section 5, Tables 2 and 3] The pooled Pearson correlations between similarity distances and raw downstream MSE/classification losses are likely confounded by target-dataset difficulty. Because E_{X,Y} is the raw MSE on target Y without per-target standardization, hard targets contribute high losses for every source; if a similarity metric correlates with target volatility (e.g., WSD depends on sample covariance), the pooled correlation can be positive even when the metric does not rank sources correctly within any target. The paper reports no within-target correlations, no per-target normalization, and no random-ranking baseline. This directly affects the main empirical conclusion that no similarity metric is consistently best, so the analysis must be re-done with target-level controls.
  2. [Section 4.4 and 5] The experimental protocol omits values for several load-bearing hyperparameters: the k in kNN membership, the quantile threshold tau, the MMD kernel bandwidth, MAD alpha/beta, DBA iteration count, Sinkhorn regularization (if used), and the fine-tuning hyperparameters for Lag-Llama and Time-MoE. Tables 2 and 3 are the only empirical evidence, and without these settings the results are not reproducible from the paper alone. The toolbox's reproducibility claim requires at least a hyperparameter table and seed information.
  3. [Section 5] The experiments use exactly one dataset pool (25 GluonTS datasets) and one sampling configuration (100 windows of length 100). The abstract and introduction claim 'diverse experimental settings,' but no variation in corpus, sampling, or normalization is reported. The claim that the findings generalize across settings is therefore unsupported. Please either add experiments with different pools/settings or temper the claim accordingly.
  4. [Section 5, Tables 2 and 3] The correlation values are point estimates derived from a single run; no confidence intervals, p-values, or repeated seeds are reported. Consequently, the paper cannot support its comparison of metrics (e.g., MAD vs WSD in OOD-TMoE) or its claim about consistency across tasks. Please provide error bars or significance testing.
minor comments (4)
  1. [First page] The copyright block reads 'Conference acronym ’XX, Woodstock, NY ©2018'; the year should match the intended publication venue.
  2. [Section 4.3] In the WSD formula, (Σ_X Σ_Y)^{1/2} is the matrix square root; please state this explicitly to avoid ambiguity.
  3. [Table 1] The 'Equal Length' column shows ✗ for MAD, but the text says MAD supports different lengths; please clarify the intended meaning of that column entry.
  4. [Reference [7]] Reference [7] is co-authored by two of the paper's authors; please add a disclosure or footnote to make the self-citation transparent.

Circularity Check

0 steps flagged · score 1.0 of 10

No circular derivation: TSDS-Toolbox compares fixed external similarity metrics against independently computed downstream losses; the only WSD self-citation is not load-bearing.

full rationale

This paper makes no first-principles derivation or prediction claim; its contribution is an engineering benchmark and an empirical comparison. Each similarity metric is a fixed formula taken from prior work (WSD from [7], MMD from [16], OT from [2], MAD from [26]), and the downstream losses are computed independently by fine-tuning Lag-Llama/Time-MoE for forecasting and by a kNN membership rule for classification. There is no fitted parameter that maps similarity scores to downstream losses, so none of the reported correlations can reduce to a fit by construction. The only self-citation is WSD [7], whose authors overlap with the present paper; however, the toolbox quotes the complete Frechet-style formula and applies it without adaptation, and the paper's conclusion that no similarity metric consistently aligns with downstream performance does not rely on WSD being correct. The classification 'downstream' score is itself a distribution-membership dissimilarity, which makes the classification evaluation conceptually close to comparing similarity measures, but it is defined independently via kNN and does not use any of the benchmarked metrics as input; this affects interpretation, not circularity. The pooled Pearson correlations in Sec. 5 could be confounded by target difficulty, but that is a statistical validity limitation, not a circularity. Thus no circular step is exhibited; the score of 1 acknowledges only the minor, non-load-bearing WSD self-citation.

Assumptions & free parameters 8 free parameters · 4 assumptions · 0 invented entities

The central claim rests on several modeling choices that are either unspecified (kernel bandwidth, kNN k, LCSS thresholds, DBA settings) or unjustified (window subsampling, Pearson correlation as the validity metric). No new entities are introduced. These choices are not fitted to the results, so the circularity burden is low, but they limit the strength of the empirical conclusions.

free parameters (8)
  • Window count and length = 100 windows, length 100
    Every dataset is sampled with replacement to 100 windows of length 100 before similarity computation; this choice is not justified and affects all downstream scores.
  • MMD kernel bandwidth = unspecified
    MMD requires a kernel; the default estimator is stated but the kernel and its bandwidth are not given, making the MMD numbers underdetermined.
  • kNN neighborhood size (k) = unspecified
    Classification evaluation uses k-NN membership distance; k is not reported.
  • Classification acceptance quantile (tau) = source-calibrated quantile threshold, value unspecified
    The one-class membership rate depends on tau, which is not reported.
  • LCSS match threshold and temporal window = default unspecified
    LCSS normalized distance depends on a Euclidean threshold and optional temporal window; neither is reported.
  • MAD alpha and beta = alpha=1, beta=0
    Defaults are chosen in the implementation; no sensitivity analysis is provided.
  • DBA iterations and initialization = unspecified
    DTW Barycenter Averaging requires iterations and initialization; these are not reported.
  • Sinkhorn regularization (if used) = unspecified
    OT supports exact and Sinkhorn approximation; no specification of which variant or regularization is used.
assumptions (4)
  • domain assumption Pearson correlation between dataset-distance scores and downstream task losses is a valid measure of similarity-metric utility.
    The entire evaluation relies on this proxy; no alternative validation or theoretical justification is provided.
  • domain assumption Sampling each dataset with replacement to 100 windows of length 100 preserves the essential structure for dataset similarity.
    The experiments use only this subsampling setting, so conclusions about metric quality are conditioned on it.
  • domain assumption Removing flat sequences and applying z-score normalization does not bias comparisons across metrics.
    Preprocessing is applied uniformly, but its effect on different metrics is not analyzed.
  • domain assumption The one-class kNN membership rule and MSE loss adequately capture downstream transfer performance.
    Classification and forecasting evaluation are reduced to these two proxies; other downstream tasks are not considered.

how reviews work

0 comments
Cite this review

Pith. "Pith review of TSDS-Toolbox: A Toolbox for Measuring Time-Series Dataset Similarity." pith.science (2026). https://pith.science/paper/MAYN43DP

@misc{pith2026260808119,
  author       = {Pith},
  title        = {Pith review of: TSDS-Toolbox: A Toolbox for Measuring Time-Series Dataset Similarity},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MAYN43DP}},
  note         = {Machine review of arXiv:2608.08119}
}
read the original abstract

The rapid advancement of artificial intelligence (AI) has significantly accelerated research in time-series analysis, particularly in forecasting, classification, and generation tasks. Recent models, especially foundation models, benefit from time-series dataset similarity due to its significant role in source dataset selection for fine-tuning. However, many existing implementations for benchmarking time-series dataset similarity methods are fragmented and difficult to extend. To address this, we present a unified framework, the Time-Series Dataset Similarity Toolbox (TSDS-Toolbox). Our work enables (1) systematic and reproducible comparisons of time-series dataset similarity methods; (2) flexible extensibility for users to add customized datasets, similarity methods, and downstream time-series tasks; and (3) consistent evaluation of both dataset-level and series-level similarity methods through integrated time-series dataset reducers. The effectiveness of TSDS-Toolbox is validated through comprehensive experiments under diverse experimental settings. Our toolbox is publicly available.

Figures

Figures reproduced from arXiv: 2608.08119 by the authors.

Figure 1
Figure 1. Overview of the TSDS-Toolbox architecture and its modules. [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Visualization results generated by TSDS-Toolbox. [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

44 extracted references · 22 canonical work pages

  1. [7]

    Measuring Time-Series Dataset Similarity using Wasserstein Distance

    Hongjie Chen, Akshay Mehra, Josh Kimball, and Ryan A Rossi. Measur- ing time-series dataset similarity using wasserstein distance.arXiv preprint arXiv:2507.22189, 2025

    work page Pith review arXiv 2025

  2. [1]

    Gluonts: Probabilistic and neural time series modeling in python.Journal of Machine Learning Research, 21(116):1–6, 2020

    Alexander Alexandrov, Konstantinos Benidis, Michael Bohlke-Schneider, Valentin Flunkert, Jan Gasthaus, Tim Januschowski, Danielle C Maddix, Syama Rangapu- ram, David Salinas, Jasper Schulz, et al. Gluonts: Probabilistic and neural time series modeling in python.Journal of Machine Learning Research, 21(116):1–6, 2020

    work page 2020

  3. [2]

    Geometric dataset distances via optimal transport.Advances in Neural Information Processing Systems, 33:21428–21439, 2020

    David Alvarez-Melis and Nicolo Fusi. Geometric dataset distances via optimal transport.Advances in Neural Information Processing Systems, 33:21428–21439, 2020

    2020

  4. [3]

    Chronos: Learning the language of time series.arXiv preprint arXiv:2403.07815, 2024

    Abdul Fatir Ansari, Lorenzo Stella, Caner Turkmen, Xiyuan Zhang, Pedro Mercado, Huibin Shen, Oleksandr Shchur, Syama Sundar Rangapuram, Sebas- tian Pineda Arango, Shubham Kapoor, et al. Chronos: Learning the language of time series.arXiv preprint arXiv:2403.07815, 2024

    arXiv 2024

  5. [4]

    wav2vec 2.0: A framework for self-supervised learning of speech representations.Advances in neural information processing systems, 33:12449–12460, 2020

    Alexei Baevski, Yuhao Zhou, Abdelrahman Mohamed, and Michael Auli. wav2vec 2.0: A framework for self-supervised learning of speech representations.Advances in neural information processing systems, 33:12449–12460, 2020

    2020

  6. [5]

    The uea multivariate time series classification archive, 2018.arXiv preprint arXiv:1811.00075, 2018

    Anthony Bagnall, Hoang Anh Dau, Jason Lines, Michael Flynn, James Large, Aaron Bostrom, Paul Southam, and Eamonn Keogh. The uea multivariate time series classification archive, 2018.arXiv preprint arXiv:1811.00075, 2018

    arXiv 2018

  7. [6]

    Recurrent neural networks for multivariate time series with missing values

    Zhengping Che, Sanjay Purushotham, Kyunghyun Cho, David Sontag, and Yan Liu. Recurrent neural networks for multivariate time series with missing values. Scientific reports, 8(1):6085, 2018

    2018

  8. [8]

    Nearest neighbor pattern classification.IEEE transactions on information theory, 13(1):21–27, 1967

    Thomas Cover and Peter Hart. Nearest neighbor pattern classification.IEEE transactions on information theory, 13(1):21–27, 1967

    1967

Show all 44 references
  1. [9]

    Soft-dtw: a differentiable loss function for time-series

    Marco Cuturi and Mathieu Blondel. Soft-dtw: a differentiable loss function for time-series. InInternational conference on machine learning, pages 894–903. PMLR, 2017

    2017

  2. [10]

    A decoder-only foundation model for time-series forecasting.arXiv preprint arXiv:2310.10688, 2023

    Abhimanyu Das, Weihao Kong, Rajat Sen, and Yichen Zhou. A decoder-only foundation model for time-series forecasting.arXiv preprint arXiv:2310.10688, 2023

    2023 arXiv

  3. [11]

    Querying and mining of time series data: Experimental comparison of representations and distance measures.Proc

    Hui Ding, Goce Trajcevski, Peter Scheuermann, Xiaoyue Wang, and Eamonn J Keogh. Querying and mining of time series data: Experimental comparison of representations and distance measures.Proc. VLDB Endow., 1(2):1542–1552, 2008

    2008

  4. [12]

    The impact of data set similarity and diversity on transfer learning success in time series forecasting.arXiv preprint arXiv:2404.06198, 2024

    Claudia Ehrig, Benedikt Sonnleitner, Ursula Neumann, Catherine Cleophas, and Germain Forestier. The impact of data set similarity and diversity on transfer learning success in time series forecasting.arXiv preprint arXiv:2404.06198, 2024

    2024 arXiv

  5. [13]

    Fast subsequence matching in time-series databases.ACM Sigmod Record, 23(2):419– 429, 1994

    Christos Faloutsos, Mudumbai Ranganathan, and Yannis Manolopoulos. Fast subsequence matching in time-series databases.ACM Sigmod Record, 23(2):419– 429, 1994

    1994

  6. [14]

    Transfer learning for time series classification

    Hassan Ismail Fawaz, Germain Forestier, Jonathan Weber, Lhassane Idoumghar, and Pierre-Alain Muller. Transfer learning for time series classification. In2018 IEEE international conference on big data (Big Data), pages 1367–1376. IEEE, 2018

    2018

  7. [15]

    Computational optimal transport with applica- tions to data sciences.Foundations and Trends®in Machine Learning, 11(5-6):355– 607, 2019

    Peyré Gabriel and Cuturi Marco. Computational optimal transport with applica- tions to data sciences.Foundations and Trends®in Machine Learning, 11(5-6):355– 607, 2019

    2019

  8. [16]

    A kernel two-sample test.The journal of machine learning research, 13(1):723–773, 2012

    Arthur Gretton, Karsten M Borgwardt, Malte J Rasch, Bernhard Schölkopf, and Alexander Smola. A kernel two-sample test.The journal of machine learning research, 13(1):723–773, 2012

    2012

  9. [17]

    Retrieval augmented time series forecasting.arXiv preprint arXiv:2505.04163, 2025

    Sungwon Han, Seungeon Lee, Meeyoung Cha, Sercan O Arik, and Jinsung Yoon. Retrieval augmented time series forecasting.arXiv preprint arXiv:2505.04163, 2025

    2025 arXiv

  10. [18]

    Multitask learning and benchmarking with clinical time series data

    Hrayr Harutyunyan, Hrant Khachatrian, David C Kale, Greg Ver Steeg, and Aram Galstyan. Multitask learning and benchmarking with clinical time series data. Scientific data, 6(1):96, 2019

    2019

  11. [19]

    Darts: User-friendly modern machine learning for time series.Journal of Machine Learning Research, 23(124):1–6, 2022

    Julien Herzen, Francesco Lässig, Samuele Giuliano Piazzetta, Thomas Neuer, Léo Tafti, Guillaume Raille, Tomas Van Pottelbergh, Marek Pasieka, Andrzej Skrodzki, Nicolas Huguenin, et al. Darts: User-friendly modern machine learning for time series.Journal of Machine Learning Res...

    2022

  12. [20]

    Gans trained by a two time-scale update rule converge to a local nash equilibrium.Advances in neural information processing systems, 30, 2017

    Martin Heusel, Hubert Ramsauer, Thomas Unterthiner, Bernhard Nessler, and Sepp Hochreiter. Gans trained by a two time-scale update rule converge to a local nash equilibrium.Advances in neural information processing systems, 30, 2017

    2017

  13. [21]

    Time-llm: Time series forecasting by reprogramming large language models

    Ming Jin, Shiyu Wang, Lintao Ma, Zhixuan Chu, James Zhang, Xiaoming Shi, Pin-Yu Chen, Yuxuan Liang, Yuan-Fang Li, Shirui Pan, et al. Time-llm: Time series forecasting by reprogramming large language models. InInternational conference on learning representations, volume 2024, p...

    2024

  14. [22]

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

    Xu Liu, Junfeng Hu, Yuan Li, Shizhe Diao, Yuxuan Liang, Bryan Hooi, and Roger Zimmermann. Unitime: A language-empowered unified model for cross-domain time series forecasting. InProceedings of the ACM Web Conference 2024, pages 4095–4106, 2024

    2024

  15. [23]

    sktime: A unified interface for machine learning with time series.arXiv preprint arXiv:1909.07872, 2019

    Markus Löning, Anthony Bagnall, Sajaysurya Ganesh, Viktor Kazakov, Jason Lines, and Franz J Király. sktime: A unified interface for machine learning with time series.arXiv preprint arXiv:1909.07872, 2019

    1909 arXiv

  16. [24]

    Climax: A foundation model for weather and climate.arXiv preprint arXiv:2301.10343, 2023

    Tung Nguyen, Johannes Brandstetter, Ashish Kapoor, Jayesh K Gupta, and Aditya Grover. Climax: A foundation model for weather and climate.arXiv preprint arXiv:2301.10343, 2023

    2023 arXiv

  17. [25]

    Kanghui Ning, Zijie Pan, Yu Liu, Yushan Jiang, James Zhang, Kashif Rasul, Ander- son Schneider, Lintao Ma, Yuriy Nevmyvaka, and Dongjin Song. Ts-rag: Retrieval- augmented generation based time series foundation models are stronger zero-shot forecaster.Advances in Neural Inform...

    2026

  18. [26]

    Match-and-deform: Time series domain adap- tation through optimal transport and temporal alignment

    François Painblanc, Laetitia Chapel, Nicolas Courty, Chloé Friguet, Charlotte Pelletier, and Romain Tavenard. Match-and-deform: Time series domain adap- tation through optimal transport and temporal alignment. InJoint European Conference on Machine Learning and Knowledge Disco...

    2023

  19. [27]

    Tsb-uad: An end-to-end benchmark suite for univariate time-series anomaly detection.Proc

    John Paparrizos, Yuhao Kang, Paul Boniol, Ruey S Tsay, Themis Palpanas, and Michael J Franklin. Tsb-uad: An end-to-end benchmark suite for univariate time-series anomaly detection.Proc. VLDB Endow., 15(8):1697–1711, 2022

    2022

  20. [28]

    Karl Pearson. Vii. note on regression and inheritance in the case of two parents. proceedings of the royal society of London, 58(347-352):240–242, 1895

  21. [29]

    Karl Pearson. Liii. on lines and planes of closest fit to systems of points in space. The London, Edinburgh, and Dublin philosophical magazine and journal of science, 2(11):559–572, 1901

    1901

  22. [30]

    A global averaging method for dynamic time warping, with applications to clustering.Pattern recognition, 44(3):678–693, 2011

    François Petitjean, Alain Ketterlin, and Pierre Gançarski. A global averaging method for dynamic time warping, with applications to clustering.Pattern recognition, 44(3):678–693, 2011

    2011

  23. [31]

    Tfb: Towards comprehensive and fair benchmarking of time series forecasting methods.arXiv preprint arXiv:2403.20150, 2024

    Xiangfei Qiu, Jilin Hu, Lekui Zhou, Xingjian Wu, Junyang Du, Buang Zhang, Chenjuan Guo, Aoying Zhou, Christian S Jensen, Zhenli Sheng, et al. Tfb: Towards comprehensive and fair benchmarking of time series forecasting methods.arXiv preprint arXiv:2403.20150, 2024

    2024 arXiv

  24. [32]

    Lag-llama: Towards foundation models for probabilistic time series forecasting.arXiv preprint arXiv:2310.08278, 2023

    Kashif Rasul, Arjun Ashok, Andrew Robert Williams, Hena Ghonia, Rishika Bhag- watkar, Arian Khorasani, Mohammad Javad Darvishi Bayazi, George Adamopou- los, Roland Riachi, Nadhir Hassen, et al. Lag-llama: Towards foundation models for probabilistic time series forecasting.arXi...

    arXiv

  25. [33]

    Dynamic programming algorithm optimization for spoken word recognition.IEEE transactions on acoustics, speech, and signal processing, 26(1):43–49, 1978

    Hiroaki Sakoe and Seibi Chiba. Dynamic programming algorithm optimization for spoken word recognition.IEEE transactions on acoustics, speech, and signal processing, 26(1):43–49, 1978

    1978

  26. [34]

    Time-moe: Billion-scale time series foundation models with mixture of experts

    Xiaoming Shi, Shiyu Wang, Yuqi Nie, Dianqi Li, Zhou Ye, Qingsong Wen, and Ming Jin. Time-moe: Billion-scale time series foundation models with mixture of experts. InInternational conference on learning representations, volume 2025, pages 34635–34667, 2025

    2025

  27. [35]

    Quantifying dataset similarity to guide transfer learning.arXiv preprint arXiv:2510.10866, 2025

    Shudong Sun and Hao Helen Zhang. Quantifying dataset similarity to guide transfer learning.arXiv preprint arXiv:2510.10866, 2025

    2025 arXiv

  28. [36]

    Tslearn, a machine learning toolkit for time series data.Journal of machine learning research, 21(118):1–6, 2020

    Romain Tavenard, Johann Faouzi, Gilles Vandewiele, Felix Divo, Guillaume An- droz, Chester Holtz, Marie Payne, Roman Yurchak, Marc Rußwurm, Kushal Kolar, et al. Tslearn, a machine learning toolkit for time series data.Journal of machine learning research, 21(118):1–6, 2020

    2020

  29. [37]

    Discovering similar multidimensional trajectories

    Michail Vlachos, George Kollios, and Dimitrios Gunopulos. Discovering similar multidimensional trajectories. InProceedings 18th international conference on data engineering, pages 673–684. IEEE, 2002

    2002

  30. [38]

    Deep time series models: A comprehensive survey and benchmark.IEEE Transactions on Pattern Analysis and Machine Intelligence, 2026

    Yuxuan Wang, Haixu Wu, Jiaxiang Dong, Yong Liu, Chen Wang, Mingsheng Long, and Jianmin Wang. Deep time series models: A comprehensive survey and benchmark.IEEE Transactions on Pattern Analysis and Machine Intelligence, 2026

    2026

  31. [39]

    Unified training of universal time series forecasting transformers

    Gerald Woo, Chenghao Liu, Akshat Kumar, Caiming Xiong, Silvio Savarese, and Doyen Sahoo. Unified training of universal time series forecasting transformers. InForty-first International Conference on Machine Learning, 2024

    2024

  32. [40]

    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. arXiv preprint arXiv:2210.02186, 2022

    2022 arXiv

  33. [41]

    Estimating time series foundation model transferability via in-context learning.arXiv preprint arXiv:2509.23695, 2025

    Qingren Yao, Ming Jin, Chengqi Zhang, Chao-Han Huck Yang, Jun Qi, and Shirui Pan. Estimating time series foundation model transferability via in-context learning.arXiv preprint arXiv:2509.23695, 2025

    2025

  34. [42]

    Stock price prediction via discovering multi-frequency trading patterns

    Liheng Zhang, Charu Aggarwal, and Guo-Jun Qi. Stock price prediction via discovering multi-frequency trading patterns. InProceedings of the 23rd ACM SIGKDD international conference on knowledge discovery and data mining, pages 2141–2149, 2017

    2017

  35. [43]

    Unified transferability metrics for time series foundation models

    Weiyang Zhang, Xinyang Chen, Xiucheng Li, Kehai Chen, Weili Guan, and Liqiang Nie. Unified transferability metrics for time series foundation models. Advances in Neural Information Processing Systems, 38:41155–41178, 2026

    2026

  36. [44]

    A similarity-based time series source dataset selection method for transfer learning

    Zongxi Zhang, Kejian Zhou, Xiaohu Wang, Ming Zhu, Jiahao Gao, Jie Mei, and Qinqing Huang. A similarity-based time series source dataset selection method for transfer learning. InJournal of Physics: Conference Series, volume 2428, page 012038. IOP Publishing, 2023

    2023

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.