Pith. sign in

REVIEW 3 major objections 5 minor 49 references

Implet: A Post-hoc Subsequence Explainer for Time Series Models

T0 review · 3 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read Implet finds the exact time segments that drive model predictions.

desk verdict A clean, well-written method paper that combines attribution maps, contiguous high-score extraction, and DTW/DBA clustering into a genuinely new combination, but the faithfulness evidence needs stronger baselines and statistics before the strong claims hold. read the letter →

arxiv 2505.08748 v1 pith:TA7BLJVI submitted 2025-05-13 cs.LG

Fanyu Meng , Ziwen Kan , Shahbaz Rezaei , Zhaodan Kong , Xin Chen , Xin Liu This is my paper · ORCID
classification cs.LG
keywords timeseriesexplainabilitypost-hocexplanationsubsequencefeatureattributioncohortshapeletsfaithfulnessevaluationclassification
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 Implet, a post-hoc explainer that condenses any per-timestep attribution map for a time series classifier into a small set of contiguous subsequences, called implets. The central claim is faithfulness: in ablation experiments on 13 benchmark datasets, removing the implets a model relies on causes substantially larger accuracy drops than removing random subsequences of equal length. Implet's score for a candidate subsequence is the sum of absolute attribution values plus a length bonus, $\sum_{i=l}^{r}|w_i| + \lambda(r-l+1)$, and extraction runs in linear time via a greedy scan. The paper also introduces Coh-Implet, which clusters implets into representative centroids using dynamic time warping, so users can see group-level explanations that remain nearly as faithful as instance-level ones. The motivation is to give deep learning time series classifiers the intuitive, shapelet-like explanations they currently lack, without sacrificing the classifiers' accuracy.

What carries the argument

The central object is the implet, a subsequence $I(l,r;x,w)$ paired with its attribution values, scored by $s(l,r)= \sum_{i=l}^{r} |w_i| + \lambda(r-l+1)$ and constrained to length between $\ell_{\min}$ and $\ell_{\max}$. The score's length bonus prevents a single high-attribution spike from being reported, while the cumulative absolute attribution ensures the segment is where the model looks. Extraction is a greedy $O(T)$ scan that finds the best end location for each start position and then skips past the chosen implet. For cohort explanations, the machinery is two-dimensional dependent Dynamic Time Warping (DTW) distance—using both feature values and attribution values—together with DTW Barycenter Averaging (DBA) to compute centroids, and the silhouette score to pick the number of clusters. The removal-based faithfulness evaluation uses a randomized polynomial interpolation to erase a subsequence smoothly, avoiding artificial discontinuities that would confound the ablation.

What would settle it

Take a synthetic time series with a known single decision-relevant motif (e.g., a handcrafted 20-step pattern embedded in noise that fully determines the class label). Train a small CNN, compute Implet with Saliency, and check whether the extracted implet matches the true motif and whether removing it drops accuracy well below random removal; if Implet systematically misses the motif or its removal drop ties random removal, the attribution-to-subsequence assumption is refuted.

Watch

Extended reading notes

Core claim

On the paper's own terms, Implet establishes that subsequence-level explanations can be both concise and faithful for state-of-the-art time series models. Given a classifier and one input, Implet takes class-specific attribution scores (from any method—saliency, Input×Gradient, DeepLIFT, LIME, SHAP, occlusion) and extracts non-overlapping subsequences whose cumulative absolute attribution plus a length term exceeds a threshold. In the faithfulness test, removing each identified implet with a smooth polynomial interpolation drops accuracy far more than removing random segments in nearly all dataset/model/attribution combinations, and clearly outperforms the ShapeletTransform baseline. Coh-Implet goes further: clustering implets by two-dimensional dependent dynamic time warping and reporting the DTW barycenter as a centroid yields explanations that, when matched to new samples by shape alone, still produce near-implet-level accuracy drops—showing the centroids capture model-relevant structure rather than dataset statistics.

Load-bearing premise

Implet assumes that feature attribution scores accurately reflect which time steps drive the model's prediction, so that a contiguous stretch of high absolute attribution marks a genuinely decision-relevant subsequence.

Editorial extensions

If this is right

  • Any black-box time series classifier—CNN, InceptionTime, or other—can now be given concise, faithful explanations: a handful of highlighted segments per class instead of a dense attribution vector.
  • Removal-ablation can become a standard, architecture-aware faithfulness test for time series explainers, and Implet passes it across most UCR benchmarks.
  • Coh-Implet provides group-level summaries that retain most of the faithfulness of instance-level implets, making it feasible to summarize model behavior over entire datasets or classes.
  • The method is attribution-agnostic, so it can be layered onto any existing saliency or perturbation explainer to make its output more human-readable.
  • On frequency-based (FordA) and event-based (Earthquakes) data, subsequence explanations run into a hard limit; the paper maps out where this explanation paradigm should and should not be expected to work.

Reading between the lines

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

  • I would expect the faithful-subsequence idea to transfer to forecasting and anomaly detection, where identifying the exact temporal window driving a prediction or alert is often more actionable than a per-timestep score; the paper does not test these settings.
  • The two-dimensional DTW clustering could be repurposed as a post-hoc shapelet discovery tool: the cohort centroids are, in effect, shapelets extracted from a trained model rather than from data statistics, which might give the best of both paradigms.
  • The smooth polynomial removal procedure is itself a reusable ablation component; future benchmarks could adopt it to reduce artifacts when testing any perturbation-based explainer on smooth or short time series.
  • A testable extension of the faithfulness claim: on longer multivariate series, generalizing implets to $2n$ dimensions (value plus attribution per channel) may yield much sparser and more precise explanations than per-channel attribution maps alone.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper introduces Implet, a post-hoc subsequence explainer for time series classifiers. Implet first computes feature attributions for a target model and then extracts contiguous subsequences whose cumulative absolute attribution plus a length bonus exceeds a threshold, subject to length bounds. A second contribution, Coh-Implet, clusters the extracted subsequences using a two-dimensional dependent DTW distance with DBA centroids and silhouette-based selection of the number of clusters, yielding cohort-level explanations. The method is evaluated on 13 UCR binary classification datasets with FCN and InceptionTime classifiers and six attribution methods. Faithfulness is measured by removing the identified subsequences, using a smooth polynomial replacement scheme, and comparing the accuracy drop with that from removing random subsequences of equal length. The main quantitative claims are that removing implets almost always causes significantly larger accuracy drops than random removal, and that cluster centroids (Coh-Implets) remain faithful when matched by shape alone.

Significance. If the claims are substantiated, Implet provides a simple and inexpensive way to turn arbitrary feature attribution maps into subsequence-level explanations, and the cohort extension is a natural step toward more concise explanations for time series. The paper is clearly written, the algorithms are transparent, and the authors release code. The cohort experiment in Section IV-C is a particularly good idea: matching cohort centroids using only the feature dimension and then measuring faithfulness provides a non-circular check that is stronger than the main ablation. However, the central faithfulness claim currently rests on a comparison against random subsequences, which is too weak to isolate the value of Implet's construction, and the empirical results are reported as point estimates without variance or statistical tests. The manuscript therefore needs additional baselines and a more rigorous experimental protocol before the main claims can be accepted.

major comments (3)
  1. [IV-B, Eq. (2)] The main faithfulness experiment compares removing implets with removing random subsequences of equal length. Because Eq. (2) defines implets as contiguous runs of high cumulative absolute attribution, the observed gap in accuracy drop is largely a consequence of the attribution scores themselves; the experiment does not test whether Implet's specific aggregation choices—contiguity, length bounds, the lambda length bonus, and the greedy endpoint selection in Algorithm 1—add explanatory value. Please add baselines that consume the same attribution scores without these choices, for example a fixed-length window centered on the maximum attribution, a non-contiguous top-k set of the same cardinality, and random contiguous windows sampled from the same length distribution as the extracted implets. The claim that Implet improves over raw attribution maps depends on showing that these alternatives are worse than Implet.
  2. [IV-B, Figs. 5-6] The reported accuracy drops are single point estimates with no error bars, confidence intervals, or statistical tests. The claim that removing implets leads to 'significantly larger' accuracy drops is therefore not supported by the presented evidence. Please report repeated runs (for example, multiple model training seeds and multiple random removal seeds), show the variance across samples or runs, and apply a paired test between removal conditions. This is especially important because Fig. 5 shows several datasets where the random and implet removals appear very close (e.g., GunPoint with InceptionTime) and concluding 'almost always' requires knowing how stable the point estimates are.
  3. [Appendix B] The evaluation protocol changes for Chinatown, Earthquakes, and FordA are introduced only after the main results in Fig. 5 are found to be unfavorable for those datasets: Chinatown switches to mean-fill removal, and Earthquakes/FordA switch to removing all implets rather than one per sample. This is a post-hoc change in the experimental design, and it is used to recover the desired conclusion. Please either pre-specify the criteria under which the alternative protocol applies, or report both the main and alternative protocols for all datasets so that the comparison remains apples-to-apples. At minimum, the paper should state explicitly that the more favorable results in Fig. 8 are obtained under a different protocol and are not directly comparable to the results for the other ten datasets.
minor comments (5)
  1. [IV-C and Fig. 6] The abbreviation for the cohort-matched subsequences is inconsistently given as both ICLS (in the text of Section IV-C) and CILS (in the caption of Fig. 6); please unify the notation.
  2. [Algorithm 1] Line 4 of Algorithm 1 checks 'w_i >= phi' but Eq. (2) accumulates absolute values |w_i|; please clarify whether the threshold in the algorithm is applied to |w_i| or to the raw attribution, and adjust the pseudocode accordingly.
  3. [Throughout] There are several typos and grammatical slips, including 'Shaplets' and 'shaplets' in Sections II and IV-B, 'Earthquaks' in the captions of Figs. 5 and 6, 'subseuqnces' in Appendix B, and 'The last two dataset differ' in Section IV-B; please proofread.
  4. [IV-B] The sentence 'This might due to GuidedBackprop zero-ing out negative gradients' is incomplete; please rephrase.
  5. [II-A] The paper claims that LASTS is the only prior post-hoc subsequence explainer; given the breadth of the time-series XAI literature, please either provide a more thorough search statement or soften the uniqueness claim.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: Implet's faithfulness claim is supported by an empirical ablation, not by construction from its definition.

full rationale

I walked the paper's derivation chain and found no circular step that reduces a claimed result to its own inputs. Implet is defined (Eq. 2) as a contiguous subsequence with high cumulative absolute attribution plus a length bonus, but the paper's central faithfulness claim is not an analytical consequence of that definition. The Section IV-B ablation is an empirical test: removing implets is compared against removing random subsequences of equal length, and this test can fail (and does fail on some datasets, as the paper reports). The paper explicitly states its load-bearing assumption that attribution methods accurately reflect model behavior, which makes the evaluation a joint test of attribution and aggregation rather than a disguised tautology. No parameter is fitted to the faithfulness labels and then renamed as a prediction. The self-citations, [26] and [48], are used only as related-work context and as a domain observation about frequency- and event-based datasets; neither carries the derivation. The evaluation's weakness—that it does not compare Implet against alternative subsequence extractors built from the same attribution scores—is a concern about experimental strength and attribution-method dependence, not a circularity. I therefore find no self-definitional, fitted-input, self-citation-load-bearing, or renaming circularity, and assign score 0.

Assumptions & free parameters 4 free parameters · 3 assumptions · 0 invented entities

The central claim depends on hand-set thresholds (lambda, phi, length bounds) and on two domain assumptions: attribution scores are faithful and the polynomial-removal probe measures causal importance. Clustering parameters kmax and Q are not reported, adding hidden choices. No new physical entities are introduced.

free parameters (4)
  • lambda (length bonus) = 0.1
    Hand-set hyperparameter in Eq. 2 that adds a length reward to the implet score; no tuning curve or sensitivity analysis is reported.
  • phi (score threshold) = 1
    Threshold in Eq. 2; set to about one standard deviation above the mean of normalized attributions, but the normalization and distribution are not fully specified.
  • length bounds lmin and lmax = lmin=3, lmax=floor(T/2)
    Chosen 'consistent with previous shapelet literature' (Section III.A); affects which subsequences can be implets.
  • clustering repeats Q and kmax = not specified
    Algorithm 2 repeats Q times and searches k up to kmax, but the values used in experiments are not reported, making the cohort clustering partly under-specified.
assumptions (3)
  • domain assumption Feature attribution methods accurately reflect model behavior, with higher absolute attribution indicating more influence.
    Stated in Section III.A as the core assumption underlying Implet; if false, implets are not model-faithful.
  • domain assumption Removing a subsequence with smooth polynomial replacement is a faithful probe of causal importance; accuracy drop is not driven by artifacts or by the model's ability to infer the label from remaining context.
    Used in Section IV.B as the faithfulness measure; the paper itself notes InceptionTime's large receptive field can overcome single-segment removal and that three datasets needed an alternative protocol.
  • domain assumption Silhouette scores based on two-dimensional dependent DTW select the right number of cohorts.
    Algorithm 2 chooses k by maximizing silhouette; no stability analysis or sensitivity check is reported.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Implet: A Post-hoc Subsequence Explainer for Time Series Models." pith.science (2026). https://pith.science/paper/TA7BLJVI

@misc{pith2026250508748,
  author       = {Pith},
  title        = {Pith review of: Implet: A Post-hoc Subsequence Explainer for Time Series Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TA7BLJVI}},
  note         = {Machine review of arXiv:2505.08748}
}
read the original abstract

Explainability in time series models is crucial for fostering trust, facilitating debugging, and ensuring interpretability in real-world applications. In this work, we introduce Implet, a novel post-hoc explainer that generates accurate and concise subsequence-level explanations for time series models. Our approach identifies critical temporal segments that significantly contribute to the model's predictions, providing enhanced interpretability beyond traditional feature-attribution methods. Based on it, we propose a cohort-based (group-level) explanation framework designed to further improve the conciseness and interpretability of our explanations. We evaluate Implet on several standard time-series classification benchmarks, demonstrating its effectiveness in improving interpretability. The code is available at https://github.com/LbzSteven/implet

Figures

Figures reproduced from arXiv: 2505.08748 by the authors.

Figure 1
Figure 1. The workflow of the proposed post-hoc subsequence [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Implet cohort explanations for the GunPoint dataset. [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Implet cohort explanations for the Chinatown dataset. [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (4 more)
Figure 5
Figure 5. Figure 5: Faithfulness evaluation of implets with different attribution methods, compared against the baseline ShapeletTransform. [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: Faithfulness evaluation comparing Implets (dashed arrows) and Implet-Centroid-Like Subsequences (ICLS, solid arrows). [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: Faithfulness evaluation comparing (1) Implets, (2) CILS using feature dimension only, and (3) CILS using both feature [PITH_FULL_IMAGE:figures/full_fig_p009_7.png]
Figure 8
Figure 8. Figure 8: Alternative faithfulness evaluation on three datasets [PITH_FULL_IMAGE:figures/full_fig_p009_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

49 extracted references · 30 canonical work pages

  1. [1]

    Inceptiontime: Finding alexnet for time series classification,

    H. Ismail Fawaz, B. Lucas, G. Forestier, C. Pelletier, D. F. Schmidt, J. Weber, G. I. Webb, L. Idoumghar, P.-A. Muller, and F. Petitjean, “Inceptiontime: Finding alexnet for time series classification,” Data Mining and Knowledge Discovery , vol. 34, no. 6, pp. 1936–1962, 2020

    1936

  2. [2]

    Grad-cam: Visual explanations from deep networks via gradient-based localization,

    R. R. Selvaraju, M. Cogswell, A. Das, R. Vedantam, D. Parikh, and D. Batra, “Grad-cam: Visual explanations from deep networks via gradient-based localization,” in Proceedings of the IEEE international conference on computer vision , 2017, pp. 618–626

    2017

  3. [3]

    Axiomatic attribution for deep networks,

    M. Sundararajan, A. Taly, and Q. Yan, “Axiomatic attribution for deep networks,” in International conference on machine learning . PMLR, 2017, pp. 3319–3328

    2017

  4. [4]

    Not just a black box: Learning important features through propagating activation differences,

    A. Shrikumar, P. Greenside, A. Shcherbina, and A. Kundaje, “Not just a black box: Learning important features through propagating activation differences,” arXiv preprint arXiv:1605.01713 , 2016

    arXiv 2016

  5. [5]

    Visualizing and understanding convo- lutional networks,

    M. D. Zeiler and R. Fergus, “Visualizing and understanding convo- lutional networks,” in Computer Vision–ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceedings, Part I 13. Springer, 2014, pp. 818–833

    2014

  6. [6]

    Explaining time series predictions with dynamic masks,

    J. Crabb ´e and M. Van Der Schaar, “Explaining time series predictions with dynamic masks,” in International conference on machine learning . PMLR, 2021, pp. 2166–2177

    work page 2021

  7. [7]

    “why should i trust you?

    M. T. Ribeiro, S. Singh, and C. Guestrin, ““why should i trust you?” explaining the predictions of any classifier,” in Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining, 2016, pp. 1135–1144

    work page 2016

  8. [8]

    A unified approach to interpreting model predictions,

    S. M. Lundberg and S.-I. Lee, “A unified approach to interpreting model predictions,” Advances in neural information processing systems, vol. 30, 2017

    2017

Show all 49 references
  1. [9]

    Attention is all you need,

    A. Vaswani, “Attention is all you need,” Advances in Neural Information Processing Systems, 2017

    2017

  2. [10]

    Rethinking attention mechanism in time series classification,

    B. Zhao, H. Xing, X. Wang, F. Song, and Z. Xiao, “Rethinking attention mechanism in time series classification,” Information Sciences, vol. 627, pp. 97–114, 2023

    2023

  3. [11]

    Attention-based lstm- cnns for time-series classification,

    Q. Du, W. Gu, L. Zhang, and S.-L. Huang, “Attention-based lstm- cnns for time-series classification,” in Proceedings of the 16th ACM conference on embedded networked sensor systems , 2018, pp. 410–411

    2018

  4. [12]

    Time is not enough: Time- frequency based explanation for time-series black-box models,

    H. Chung, S. Jo, Y . Kwon, and E. Choi, “Time is not enough: Time- frequency based explanation for time-series black-box models,” in Proceedings of the 33rd ACM International Conference on Information and Knowledge Management , 2024, pp. 394–403

    2024

  5. [13]

    Explaining time series clas- sifiers through meaningful perturbation and optimisation,

    H. Meng, C. Wagner, and I. Triguero, “Explaining time series clas- sifiers through meaningful perturbation and optimisation,” Information Sciences, vol. 645, p. 119334, 2023

    2023

  6. [14]

    Counterfactual explanations without opening the black box: Automated decisions and the gdpr,

    S. Wachter, B. Mittelstadt, and C. Russell, “Counterfactual explanations without opening the black box: Automated decisions and the gdpr,”

  7. [15]

    This looks like that: deep learning for interpretable image recognition,

    C. Chen, O. Li, D. Tao, A. Barnett, C. Rudin, and J. K. Su, “This looks like that: deep learning for interpretable image recognition,” Advances in neural information processing systems , vol. 32, 2019

    2019

  8. [16]

    Instance-based counterfactual explanations for time series classification,

    E. Delaney, D. Greene, and M. T. Keane, “Instance-based counterfactual explanations for time series classification,” in International conference on case-based reasoning . Springer, 2021, pp. 32–47

    2021

  9. [17]

    Counterfactual ex- planations for multivariate time series,

    E. Ates, B. Aksar, V . J. Leung, and A. K. Coskun, “Counterfactual ex- planations for multivariate time series,” in 2021 international conference on applied artificial intelligence (ICAPAI) . IEEE, 2021, pp. 1–8

    2021

  10. [18]

    Time series shapelets: a novel technique that allows accurate, interpretable and fast classification,

    L. Ye and E. Keogh, “Time series shapelets: a novel technique that allows accurate, interpretable and fast classification,” Data mining and knowledge discovery, vol. 22, pp. 149–182, 2011

    2011

  11. [19]

    A shapelet transform for time series classification,

    J. Lines, L. M. Davis, J. Hills, and A. Bagnall, “A shapelet transform for time series classification,” in Proceedings of the 18th ACM SIGKDD international conference on Knowledge discovery and data mining , 2012, pp. 289–297

    2012

  12. [20]

    Early classification of multivariate temporal observations by extraction of interpretable shapelets,

    M. F. Ghalwash and Z. Obradovic, “Early classification of multivariate temporal observations by extraction of interpretable shapelets,” BMC bioinformatics, vol. 13, pp. 1–12, 2012

    2012

  13. [21]

    Learning dtw-shapelets for time-series classification,

    M. Shah, J. Grabocka, N. Schilling, M. Wistuba, and L. Schmidt- Thieme, “Learning dtw-shapelets for time-series classification,” in Pro- ceedings of the 3rd IKDD Conference on Data Science, 2016 , 2016, pp. 1–8

    2016

  14. [22]

    Learn- ing time-series shapelets,

    J. Grabocka, N. Schilling, M. Wistuba, and L. Schmidt-Thieme, “Learn- ing time-series shapelets,” in Proceedings of the 20th ACM SIGKDD international conference on Knowledge discovery and data mining , 2014, pp. 392–401

    2014

  15. [23]

    Explaining any time series classifier,

    R. Guidotti, A. Monreale, F. Spinnato, D. Pedreschi, and F. Giannotti, “Explaining any time series classifier,” in 2020 IEEE second interna- tional conference on cognitive machine intelligence (CogMI) . IEEE, 2020, pp. 167–176

    2020

  16. [24]

    Glocal explanations of expected goal models in soccer,

    M. Cavus, A. Stando, and P. Biecek, “Glocal explanations of expected goal models in soccer,” arXiv preprint arXiv:2308.15559 , 2023

    2023 arXiv

  17. [25]

    Vine: Visualizing statistical interactions in black box models,

    M. Britton, “Vine: Visualizing statistical interactions in black box models,” arXiv preprint arXiv:1904.00561 , 2019

    1904 arXiv

  18. [26]

    Cohex: A generalized framework for cohort explanation,

    F. Meng, X. Liu, Z. Kong, and X. Chen, “Cohex: A generalized framework for cohort explanation,” 2024. [Online]. Available: https: //arxiv.org/abs/2410.13190

    2024 arXiv

  19. [27]

    Repid: Regional effect plots with implicit interaction detection,

    J. Herbinger, B. Bischl, and G. Casalicchio, “Repid: Regional effect plots with implicit interaction detection,” in International Conference on Artificial Intelligence and Statistics . PMLR, 2022, pp. 10 209–10 233

    2022

  20. [28]

    Model-agnostic feature importance and effects with dependent features: a conditional subgroup approach,

    C. Molnar, G. K ¨onig, B. Bischl, and G. Casalicchio, “Model-agnostic feature importance and effects with dependent features: a conditional subgroup approach,” Data Mining and Knowledge Discovery , pp. 1–39, 2023

    2023

  21. [29]

    Decomposing global feature effects based on feature interactions,

    J. Herbinger, B. Bischl, and G. Casalicchio, “Decomposing global feature effects based on feature interactions,” arXiv preprint arXiv:2306.00541, 2023

    2023 arXiv

  22. [30]

    Global aggrega- tions of local explanations for black box models,

    I. Van Der Linden, H. Haned, and E. Kanoulas, “Global aggrega- tions of local explanations for black box models,” arXiv preprint arXiv:1907.03039, 2019

    1907 arXiv

  23. [31]

    Generalizing dtw to the multi-dimensional case requires an adaptive approach,

    M. Shokoohi-Yekta, B. Hu, H. Jin, J. Wang, and E. Keogh, “Generalizing dtw to the multi-dimensional case requires an adaptive approach,” Data mining and knowledge discovery , vol. 31, pp. 1–31, 2017

    2017

  24. [32]

    A global averaging method for dynamic time warping, with applications to clustering,

    F. Petitjean, A. Ketterlin, and P. Ganc ¸arski, “A global averaging method for dynamic time warping, with applications to clustering,” Pattern recognition, vol. 44, no. 3, pp. 678–693, 2011

    2011

  25. [33]

    Lstm fully convolu- tional networks for time series classification,

    F. Karim, S. Majumdar, H. Darabi, and S. Chen, “Lstm fully convolu- tional networks for time series classification,” IEEE access, vol. 6, pp. 1662–1669, 2017

    2017

  26. [34]

    Learning important features through propagating activation differences,

    A. Shrikumar, P. Greenside, and A. Kundaje, “Learning important features through propagating activation differences,” in International conference on machine learning . PMlR, 2017, pp. 3145–3153

    2017

  27. [35]

    Deep inside convolutional networks: Visualising image classification models and saliency maps,

    K. Simonyan, A. Vedaldi, and A. Zisserman, “Deep inside convolutional networks: Visualising image classification models and saliency maps,” arXiv preprint arXiv:1312.6034 , 2013

    2013 arXiv

  28. [36]

    tsai - a state-of-the-art deep learning library for time series and sequential data,

    I. Oguiza, “tsai - a state-of-the-art deep learning library for time series and sequential data,” Github, 2023. [Online]. Available: https://github.com/timeseriesAI/tsai

    2023

  29. [37]

    Captum: A unified and generic model inter- pretability library for pytorch,

    N. Kokhlikyan, V . Miglani, M. Martin, E. Wang, B. Alsallakh, J. Reynolds, A. Melnikov, N. Kliushkina, C. Araya, S. Yan, and O. Reblitz-Richardson, “Captum: A unified and generic model inter- pretability library for pytorch,” 2020

    2020

  30. [38]

    pyts: A python package for time series classification,

    J. Faouzi and H. Janati, “pyts: A python package for time series classification,” Journal of Machine Learning Research , vol. 21, no. 46, pp. 1–6, 2020. [Online]. Available: http://jmlr.org/papers/v21/19-763. html

    2020

  31. [39]

    DTAIDistance,

    W. Meert, K. Hendrickx, T. Van Craenendonck, P. Robberechts, H. Blockeel, and J. Davis, “DTAIDistance,” Aug. 2020. [Online]. Available: https://github.com/wannesm/dtaidistance

    2020

  32. [40]

    Everything you know about dynamic time warping is wrong,

    C. A. Ratanamahatana and E. Keogh, “Everything you know about dynamic time warping is wrong,” in Third workshop on mining temporal and sequential data , vol. 32. Citeseer, 2004

    2004

  33. [41]

    Forecasting time series subject to multiple structural breaks,

    M. H. Pesaran, D. Pettenuzzo, and A. Timmermann, “Forecasting time series subject to multiple structural breaks,” The Review of Economic Studies, vol. 73, no. 4, pp. 1057–1084, 2006

    2006

  34. [42]

    Does modeling a structural break improve forecast accuracy?

    T. Boot and A. Pick, “Does modeling a structural break improve forecast accuracy?” Journal of Econometrics , vol. 215, no. 1, pp. 35–59, 2020

    2020

  35. [43]

    Don’t get me wrong: How to apply deep visual interpre- tations to time series,

    C. Loeffler, W.-C. Lai, B. Eskofier, D. Zanca, L. Schmidt, and C. Mutschler, “Don’t get me wrong: How to apply deep visual interpre- tations to time series,” arXiv preprint arXiv:2203.07861 , 2022

    2022 arXiv

  36. [44]

    Towards a rigorous evaluation of xai methods on time series,

    U. Schlegel, H. Arnout, M. El-Assady, D. Oelke, and D. A. Keim, “Towards a rigorous evaluation of xai methods on time series,” in 2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW). IEEE, 2019, pp. 4197–4201

    2019

  37. [45]

    Learning saliency maps to explain deep time series classifiers,

    P. S. Parvatharaju, R. Doddaiah, T. Hartvigsen, and E. A. Rundensteiner, “Learning saliency maps to explain deep time series classifiers,” in Proceedings of the 30th ACM international conference on information & knowledge management , 2021, pp. 1406–1415

    2021

  38. [46]

    The ucr time series archive,

    H. A. Dau, A. Bagnall, K. Kamgar, C.-C. M. Yeh, Y . Zhu, S. Gharghabi, C. A. Ratanamahatana, and E. Keogh, “The ucr time series archive,” IEEE/CAA Journal of Automatica Sinica , vol. 6, no. 6, pp. 1293–1305, 2019

    2019

  39. [47]

    Sanity checks for saliency maps,

    J. Adebayo, J. Gilmer, M. Muelly, I. Goodfellow, M. Hardt, and B. Kim, “Sanity checks for saliency maps,” Advances in neural information processing systems, vol. 31, 2018

    2018

  40. [48]

    Explanation space: A new perspective into time series interpretability,

    S. Rezaei and X. Liu, “Explanation space: A new perspective into time series interpretability,” arXiv preprint arXiv:2409.01354 , 2024

    2024 arXiv

  41. [2018]

    Available: https://arxiv.org/abs/1711.00399

    [Online]. Available: https://arxiv.org/abs/1711.00399

    arXiv

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.