Pith. sign in

REVIEW 4 major objections 4 minor 37 references

Active Learning for Multiple Change Point Detection in Non-stationary Time Series with Deep Gaussian Processes

T0 review · 4 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read This paper argues that multiple change points in non-stationary time series can be detected accurately and cheaply by actively sampling where a deep Gaussian process's local spectrum changes, and it backs this with a new acquisition…

desk verdict The core algorithm is sensible and new, but the real-world F1 comparison is likely unfair to baselines because the evaluation margin is inflated, undermining the headline accuracy claim. read the letter →

arxiv 2505.20452 v1 pith:ZZZEH3FK submitted 2025-05-26 cs.LG

Hao Zhao , Rong Pan This is my paper · ORCID
classification cs.LG
keywords changepointdetectiondeepGaussianprocessactivelearningspectralanalysisnon-stationarytimeseriesacquisitionfunctionuncertaintyquantification
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

Change point detection usually assumes a specific kind of change, but real non-stationary time series can shift in mean, variance, trend, periodicity, or a mix of these. This paper tries to establish that one algorithm can handle all these change types by looking for changes in the local frequency content of a deep Gaussian process fit to the data. The algorithm, called DGP-Spectral-AL, uses an acquisition function that balances a spectral-change score with spectral uncertainty, so it spends its limited measurement budget where changes are likely and where the model is unsure. The authors claim this beats a derivative-aware active-learning baseline on all fifteen simulated change patterns and matches or exceeds full-data baselines on five real-world datasets while sampling only a fraction of the points. If true, the payoff is cheaper, more flexible change point detection for expensive measurement settings such as drilling logs, pollution monitoring, and materials experiments.

What carries the argument

The load-bearing object is the Spectral Change Detection Metric (SCDM), defined as $SCDM = SMC + SGD\cdot(1-SCM)/2$, which combines spectral mean change, spectral correlation mapper, and spectral gradient distance to catch both abrupt and gradual spectral changes while staying insensitive to uniform amplitude scaling. This metric is computed on the sliding-window Fourier transform of the DGP's predicted mean, and it is paired with a spectral uncertainty term $SU$ obtained by Monte Carlo sampling of the DGP posterior and averaging the variance of the resulting spectra. The acquisition function $a(x)=\beta\cdot SCDM(x)+(1-\beta)\cdot SU(x)$ selects the next sampling locations, with $\beta$ trading exploitation of detected spectral changes against exploration of high-uncertainty regions. The machinery works because the DGP's hierarchical warping smooths noise and models non-stationarity, so the spectral representation of its predictive mean carries the change signature that the thresholding step in Algorithm 1 converts into a set of change point estimates.

What would settle it

Recompute the F1 scores of the baseline methods with the same tolerance window $M=\delta$ used by the proposed method; if the baseline scores from the benchmark report change materially, the claim that the method beats existing techniques on real-world data is unsupported.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central claim is that change points in non-stationary time series can be localized by treating the mean prediction of a Deep Gaussian Process as a signal, transforming it with a sliding-window Fourier transform, and actively querying new points where the local spectrum is changing and the spectral estimate is uncertain. The change evidence is summarized by a Spectral Change Detection Metric, $SCDM = SMC + SGD\cdot(1-SCM)/2$, where SMC measures shifts in average spectral power, SCM measures correlation of spectral shapes, and SGD measures spectral gradient distance; this is combined with Monte-Carlo spectral uncertainty into an acquisition function $a(x)=\beta\cdot SCDM(x)+(1-\beta)\cdot SU(x)$. New samples are chosen at the maximizer, the DGP is refit, and after the active-learning budget is spent, change points are extracted by thresholding SCDM and suppressing a window around each detected peak. The paper reports that this method beats the DACD baseline in RMSE on every simulated pattern and reaches F1 scores up to 0.89 on real-world data while training on 30 initial points plus 10 batches, whereas the published baselines used the full datasets.

Load-bearing premise

The real-world comparison assumes that the tolerance window used to count a detection as correct (the paper sets it equal to its suppression interval) is the same tolerance the published baselines used; if those tolerances differ, the F1 numbers are not directly comparable.

Editorial extensions

If this is right

  • For simulated patterns including stratified, cycle, trend, systematic, shift, and their pairwise combinations, the method yields lower RMSE than DACD across all fifteen patterns.
  • On real-world datasets (Occupancy, Apple, Run log, Bee dance, Well log), the best $\beta$ setting reaches F1 up to 0.8889 on Run log and 0.8571 on Well log, trained on far fewer points than the full-data baselines.
  • The method works with both Matérn 5/2 and RBF kernels, and the $\beta$ parameter provides a meaningful exploration-exploitation dial: simpler shifts favor exploitation (high $\beta$), while periodic and high-variability patterns favor exploration (low $\beta$).
  • Because the selected samples concentrate around true change points as active learning proceeds, the method reduces measurement cost without sacrificing detection accuracy.

Reading between the lines

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

  • The SCDM formulation is not tied to short-time Fourier transforms; applying the same acquisition idea to wavelet or evolutionary spectral representations could extend the method to signals where a fixed window size is a poor match for the local dynamics.
  • The paper fixes $\beta$ for the whole run; an adaptive schedule that starts with exploration and shifts toward exploitation as uncertainty shrinks could outperform the best fixed value without needing a grid search.
  • Because the uncertainty term is computed from DGP posterior samples, the method inherits any miscalibration in the DGP; a diagnostic that checks whether spectral variances match empirical error could make the acquisition function more reliable.
  • The F1 comparison would be airtight if the evaluation margin were explicitly matched to the margin used by the baseline benchmark; the paper sets the margin equal to its suppression interval, which is a sensible choice but leaves this matching as an assumption.
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 proposes an offline multiple change point detection method, DGP-Spectral-AL, that combines deep Gaussian processes (DGPs), short-time Fourier transform spectral analysis, and active learning. The method first fits a DGP to a small initial set of points, converts the predicted mean and Monte Carlo sample paths into spectral representations, defines a Spectral Change Detection Metric (SCDM, Eq. 15) that combines spectral mean change and a weighted spectral gradient, and then selects new sampling locations by maximizing an acquisition function (Eq. 18) that balances SCDM with spectral uncertainty. Change points are finally extracted by thresholding SCDM with a suppression interval (Algorithm 1). Experiments compare the method against the authors' earlier DACD baseline on synthetic patterns and against several published change point detection methods on five real-world datasets.

Significance. If the claims were fully supported, the paper would offer a plausible and useful extension of active learning to change point detection under expensive data acquisition, with a spectral representation that is not tied to a single change pattern. The paper includes a sensitivity analysis and uses public benchmark datasets, which are strengths. However, the current evaluation does not substantiate the central claims: the real-world F1 comparison against baselines from [29] is likely invalid because the evaluation margin is inflated, the simulation claim of consistent superiority over DACD is contradicted by one of the reported pattern rows, and no uncertainty quantification is provided for any of the headline numbers. The algorithmic framework is interesting, but the evidence as presented is not yet convincing.

major comments (4)
  1. [Section 4.3, Section 5.2.2, Table 3] The F1 comparison against baselines from [29] is not valid as reported. Section 4.3 states that 'the margin of error M is set to be equal to the suppression interval δ', and Section 5.2.2 sets δ=10 for Occupancy, Run log, and Well log and δ=20 for Apple and Bee dance. The TCPD benchmark protocol in [29] uses a fixed margin (5 samples in the standard evaluation). Since the baseline F1 scores in Table 3 are taken from [29] rather than recomputed with the same margin, the proposed method is allowed a tolerance of 10 or 20 samples while the baselines are held to 5 samples. This systematically inflates the proposed method's F1 scores. The claim in the abstract that the method 'outperforms existing techniques in terms of detection accuracy' on real-world data is therefore not supported unless all baselines are recomputed under a common margin.
  2. [Section 5.1.2, Table 2] The statement that 'our method consistently outperforms DACD in RMSE across all patterns' is contradicted by the Trend (TP) row of Table 2. DACD achieves RMSE 0.5, while the proposed method's best configuration in that row is 0.4 (Matérn, A=15, β=0.5) and most configurations are substantially worse (e.g., 2.1, 3.2, 5.1, 7.1). The claim needs at least to define what 'consistently' means across the 15 settings of β, kernel, and window size, and the current table does not support a blanket statement.
  3. [Section 5.1.2, Table 2 and Table 3] No standard deviations, confidence intervals, or significance tests are reported anywhere in the experimental sections. Section 5.1.2 says each simulation was repeated 10 times, but Table 2 reports only the average RMSE, and Table 3 reports single F1 values. Given the large variation across β, kernel, and window size (e.g., Well log F1 ranges from 0.1667 to 0.8571 across configurations), the reported differences—such as the 0.4 vs 0.5 RMSE for the Trend pattern—cannot be judged as meaningful. The experimental section needs at least error bars or a statistical comparison to support the claims of superiority.
  4. [Table 3, Section 5.2.2] The paper includes random and uniform sampling ablations in Table 3, but the text does not analyze them, and in several cases the ablations are competitive with or better than the proposed acquisition function. For example, for Well log with the RBF kernel, uniform sampling achieves F1 0.6400 while the best proposed RBF configuration achieves 0.6250; for Occupancy with the RBF kernel, uniform sampling (0.5882) beats several AF settings. Since one of the central contributions is the active learning acquisition function, the paper should explicitly compare the AF against these ablations and discuss where and why active selection helps or fails. Without this, the 'sampling efficiency' claim is not established.
minor comments (4)
  1. [Section 4.3, Eq. (22)] The definition of recall in Eq. (22) is unnecessarily convoluted: writing R as an average over individual ground-truth points T_k with |T_k|=1 is equivalent to |TP(T,C)|/|T| but obscures the standard definition. Please simplify.
  2. [Section 5.2.2] The sentence 'For both datasets, window sizes and estimated distances were chosen based on domain knowledge' is inaccurate because five datasets are used, not two. It should say 'For each dataset' or 'For all datasets'.
  3. [Section 4.3 and Section 5.2.3] The detection threshold b in Algorithm 1 is never specified for any experiment. Section 5.2.3 suggests a percentile-based threshold (e.g., top 5%) as a possibility, but the text does not state which threshold was actually used to produce Tables 2 and 3. This should be reported for reproducibility.
  4. [General] There are several typos and grammatical errors, including 'it beneficial' in the Introduction, 'intruduce' in Section 3.2, 'Appedix A' in Section 5.1.2, and 'PryzlewiczV' in reference [15]. These should be corrected.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the only self-citation is a non-load-bearing baseline comparison, and the spectral-AL pipeline is evaluated against external ground truth.

full rationale

The paper's central claim is an empirical comparison of an algorithm, not a theorem or a first-principles derivation. The proposed pipeline is self-contained: a DGP is fit to the observed data, spectral representations are computed from DGP mean predictions, the acquisition function combines SCDM with MC spectral uncertainty, and change points are estimated by thresholding SCDM in Algorithm 1. No input quantity is defined in terms of the target change points, and no equation reduces to its own output by construction. The only self-citation, DACD [35], is used as a comparison baseline and described in the appendix; it does not justify the proposed method's assumptions, nor does it exclude alternatives. The external benchmark [29] supplies datasets and baseline F1 values, so the real-data comparison has independent content. The choice M = delta in Section 4.3 may affect comparability with the baselines' F1 margins, and this is a legitimate evaluation-protocol concern, but it is not circular: the margin is not fitted to force the proposed method's detections to match ground truth, and the F1 values are not implied by the algorithm's equations. No fitted parameter is renamed as a prediction, and no self-citation carries the load of the argument. Accordingly, the circularity score is 2, reflecting only a minor, non-load-bearing self-citation.

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

The method introduces a new acquisition function and metric but rests on standard DGP machinery and on assumptions that spectral features of the DGP mean indicate changes.

free parameters (6)
  • beta (exploration-exploitation weight) = Varied over {0, 0.25, 0.5, 0.75, 1}; best per dataset (e.g., 0.25 for Well log)
    Controls the balance between SCDM and spectral uncertainty in Eq. 18; no principled selection rule is given.
  • STFT window size A = A=20 for Apple, Bee dance; A=10 for Occupancy, Run log, Well log; sensitivity analysis in Table 4
    Chosen per dataset based on domain knowledge; performance is sensitive to A.
  • Suppression interval delta = delta=20 for Apple, Bee dance; delta=10 for others; sensitivity analysis in Table 4
    Used in Algorithm 1 to suppress nearby detections; also sets the F1 margin M.
  • Detection threshold b = Not specified; example suggests top 5% percentile
    User-defined; no procedure given for selecting it in practice.
  • Number of inducing points M = Not specified
    DGP variational inference requires M; the paper does not state this value.
  • MC sample size S = Not specified
    Number of posterior draws for spectral uncertainty in Eqs. 16-17.
assumptions (4)
  • standard math Deep Gaussian process posterior and variational inference as formulated in Eqs. (3)-(8)
    Assumed correctness of established DGP inference from [20].
  • domain assumption DGP predictive mean is a faithful representation of the underlying signal for spectral change detection
    Central modeling assumption in Section 4.1; the white noise test is used as a sanity check but does not guarantee spectral fidelity.
  • domain assumption Sliding-window Fourier transform of the predictive mean captures change points
    The paper assumes spectral changes correspond to structural changes; no proof is offered.
  • ad hoc to paper SCDM = SMC + SGD*(1-SCM)/2 is a suitable change indicator
    Eq. 15 is a direct extension of DISS from [30] with SMC added; no derivation.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Active Learning for Multiple Change Point Detection in Non-stationary Time Series with Deep Gaussian Processes." pith.science (2026). https://pith.science/paper/ZZZEH3FK

@misc{pith2026250520452,
  author       = {Pith},
  title        = {Pith review of: Active Learning for Multiple Change Point Detection in Non-stationary Time Series with Deep Gaussian Processes},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ZZZEH3FK}},
  note         = {Machine review of arXiv:2505.20452}
}
read the original abstract

Multiple change point (MCP) detection in non-stationary time series is challenging due to the variety of underlying patterns. To address these challenges, we propose a novel algorithm that integrates Active Learning (AL) with Deep Gaussian Processes (DGPs) for robust MCP detection. Our method leverages spectral analysis to identify potential changes and employs AL to strategically select new sampling points for improved efficiency. By incorporating the modeling flexibility of DGPs with the change-identification capabilities of spectral methods, our approach adapts to diverse spectral change behaviors and effectively localizes multiple change points. Experiments on both simulated and real-world data demonstrate that our method outperforms existing techniques in terms of detection accuracy and sampling efficiency for non-stationary time series.

Figures

Figures reproduced from arXiv: 2505.20452 by the authors.

Figure 1
Figure 1. Examples of Change Patterns in Stochastic Time Series. Each subplot illustrates a unique structural change occurring [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Architecture of a DGP. Each layer 𝐿 corresponds to a GP𝑙 with latent function 𝑓 𝑙 and inducing variables Z 𝑙 , U 𝑙 . The input to each layer is the output from the previous layer, and the model represents a nested composition F 𝑙 = 𝑓 𝑙 (𝑓 𝑙−1 (. . . 𝑓 1 (X))). with 𝐿 layers can be written as [20]: 𝑝(Y, {F 𝑙 ,U 𝑙 } 𝐿 𝑙=1 ) = Ö𝑁 𝑖=1 𝑝(𝑦𝑖 |𝑓 𝐿 𝑖 )× Ö𝐿 𝑙=1 𝑝(F 𝑙 |U 𝑙 ; F 𝑙−1 , Z 𝑙−1 )𝑝(U 𝑙 |Z 𝑙−1 ), (6) where Z 𝑙−1 deno… view at source ↗
Figure 3
Figure 3. Flowchart of Proposed Method. The methodology is achieved via three main steps: (1) DGP modeling: fit the non [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: DGP mean and variance predictions for the Well [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

37 extracted references · 35 canonical work pages

  1. [29]

    Gerrit JJ Van den Burg and Christopher KI Williams. 2020. An evaluation of change point detection algorithms.arXiv preprint arXiv:2003.06222(2020)

    arXiv 2020

  2. [1]

    Ryan Prescott Adams and David JC MacKay. 2007. Bayesian online changepoint detection.arXiv preprint arXiv:0710.3742(2007)

    arXiv 2007

  3. [2]

    Matthias Bitzer, Mona Meister, and Christoph Zimmer. 2023. Hierarchical- hyperplane kernels for actively learning gaussian process models of nonsta- tionary systems. InInternational Conference on Artificial Intelligence and Statistics. PMLR, 7897–7912

    work page 2023

  4. [3]

    Annie S Booth, S Ashwin Renganathan, and Robert B Gramacy. 2023. Contour Location for Reliability in Airfoil Simulation Experiments using Deep Gaussian Processes.arXiv preprint arXiv:2308.04420(2023)

    work page Pith review arXiv 2023

  5. [4]

    Edoardo Caldarelli, Philippe Wenk, Stefan Bauer, and Andreas Krause. 2022. Adaptive gaussian process change point detection. InInternational Conference on Machine Learning. PMLR, 2542–2571

    work page 2022

  6. [5]

    Alessandro Casini and Pierre Perron. 2024. Change-point analysis of time series with evolutionary spectra.Journal of Econometrics242, 2 (2024), 105811

    work page 2024

  7. [6]

    Haeran Cho and Piotr Fryzlewicz. 2015. Multiple-change-point detection for high dimensional time series via sparsified binary segmentation.Journal of the Royal Statistical Society Series B: Statistical Methodology77, 2 (2015), 475–507

    work page 2015

  8. [7]

    Andreas Damianou and Neil D Lawrence. 2013. Deep gaussian processes. In Artificial intelligence and statistics. PMLR, 207–215

    work page 2013

Show all 37 references
  1. [8]

    Matthew M Dunlop, Mark A Girolami, Andrew M Stuart, and Aretha L Teckentrup

  2. [9]

    Aleksandra Grzesiek, Radosław Zimroz, Paweł Śliwiński, Norbert Gomolla, and Agnieszka Wyłomańska. 2021. A Method for Structure Breaking Point Detection in Engine Oil Pressure Data.Energies14, 17 (2021), 5496

    2021

  3. [10]

    Muktesh Gupta, Rajesh Wadhvani, and Akhtar Rasool. 2022. Real-time change- point detection: A deep neural network-based adaptive approach for detecting changes in multivariate time series data.Expert Systems with Applications209 (2022), 118260

    2022

  4. [11]

    Zaid Harchaoui, Eric Moulines, and Francis Bach. 2008. Kernel change-point analysis.Advances in neural information processing systems21 (2008)

    2008

  5. [12]

    Shogo Hayashi, Yoshinobu Kawahara, and Hisashi Kashima. 2019. Active change- point detection. InAsian Conference on Machine Learning. PMLR, 1017–1032

    2019

  6. [13]

    Hamed Heydari and Sayyed Bagher Fatemi Nasrabadi. 2023. Remote sensing change detection: a comparative study of spectral distances.Geocarto Interna- tional38, 1 (2023), 2248059

    2023

  7. [14]

    Jeremias Knoblauch, Jack E Jewson, and Theodoros Damoulas. 2018. Doubly Ro- bust Bayesian Inference for Non-Stationary Streaming Data with𝛽-Divergences. Advances in Neural Information Processing Systems31 (2018)

    2018

  8. [15]

    Karolos K Korkas and Piotr PryzlewiczV. 2017. Multiple change-point detection for non-stationary time series using wild binary segmentation.Statistica Sinica (2017), 287–311

    2017

  9. [16]

    David S Matteson and Nicholas A James. 2014. A nonparametric approach for multiple change point analysis of multivariate data.J. Amer. Statist. Assoc.109, 505 (2014), 334–345

    2014

  10. [17]

    Julius Oelsmann, Marcello Passaro, Laura Sánchez, Denise Dettmering, Christian Schwatke, and Florian Seitz. 2022. Bayesian modelling of piecewise trends and discontinuities to improve the estimation of coastal vertical land motion: DiscoTimeS: a method to detect change points ...

    2022

  11. [18]

    Rong Pan and Steven E Rigdon. 2012. A Bayesian approach to change point estimation in multivariate SPC.Journal of Quality Technology44, 3 (2012), 231– 248

    2012

  12. [19]

    Karl Ropkins and James E Tate. 2021. Early observations on the impact of the COVID-19 lockdown on air quality trends across the UK.Science of the Total Environment754 (2021), 142374

    2021

  13. [20]

    Hugh Salimbeni and Marc Deisenroth. 2017. Doubly stochastic variational in- ference for deep Gaussian processes.Advances in neural information processing systems30 (2017)

    2017

  14. [21]

    Annie Sauer, Andrew Cooper, and Robert B Gramacy. 2023. Vecchia-approximated deep Gaussian processes for computer experiments.Journal of Computational and Graphical Statistics32, 3 (2023), 824–837

    2023

  15. [22]

    Annie Sauer, Robert B Gramacy, and David Higdon. 2023. Active learning for deep Gaussian process surrogates.Technometrics65, 1 (2023), 4–18

    2023

  16. [23]

    Evelien Schat, Francis Tuerlinckx, Arnout C Smit, Bart De Ketelaere, and Eva Ceulemans. 2023. Detecting mean changes in experience sampling data in real time: A comparison of univariate and multivariate statistical process control methods.Psychological Methods28, 6 (2023), 1335

    2023

  17. [24]

    Günter Schiepek, Helmut Schöller, Giulio de Felice, Sune Vork Steffensen, Marie Skaalum Bloch, Clemens Fartacek, Wolfgang Aichhorn, and Kathrin Viol

  18. [25]

    Zhenyao Shen, Xiaoshu Hou, Wen Li, Guzhanuer Aini, Lei Chen, and Yongwei Gong. 2015. Impact of landscape pattern at multiple spatial scales on water quality: A case study in a typical urbanised watershed in China.Ecological Indicators48 (2015), 417–427

    2015

  19. [26]

    Robert M Steward, Steven E Rigdon, and Rong Pan. 2016. A bayesian approach to diagnostics for multivariate control charts.Journal of Quality Technology48, 4 , , Hao Zhao and Rong Pan (2016), 303–325

    2016

  20. [27]

    Saowanit Sukparungsee, Yupaporn Areepong, and Rattikarn Taboran. 2020. Ex- ponentially weighted moving average—Moving average charts for monitoring the process mean.Plos one15, 2 (2020), e0228208

    2020

  21. [28]

    Reenu Toodesh, Sandra Verhagen, and Anastasia Dagla. 2021. Prediction of changes in seafloor depths based on time series of bathymetry observations: Dutch north sea case.Journal of Marine Science and Engineering9, 9 (2021), 931

    2021

  22. [30]

    Li Yan, Wang Xia, Zhan Zhao, and Yanran Wang. 2018. A novel approach to unsupervised change detection based on hybrid spectral difference.Remote Sensing10, 6 (2018), 841

    2018

  23. [31]

    Xiaofeng Yu and Ying Cheng. 2022. A comprehensive review and compari- son of CUSUM and change-point-analysis methods to detect test speededness. Multivariate Behavioral Research57, 1 (2022), 112–133

    2022

  24. [32]

    Min Zhang and Wenzhong Shi. 2020. A feature difference convolutional neural network-based change detection method.IEEE Transactions on Geoscience and Remote Sensing58, 10 (2020), 7232–7246

    2020

  25. [33]

    Minghe Zhang, Liyan Xie, and Yao Xie. 2023. Spectral CUSUM for Online Network Structure Change Detection.IEEE Transactions on Information Theory69, 7 (2023), 4691–4707

    2023

  26. [34]

    Yangjian Zhang, Li Wang, Quan Zhou, Feng Tang, Bo Zhang, Ni Huang, and Biswajit Nath. 2022. Continuous Change Detection and Classification—Spectral Trajectory Breakpoint Recognition for Forest Monitoring.Land11, 4 (2022), 504

    2022

  27. [35]

    Hao Zhao and Rong Pan. 2023. Active Learning for Abrupt Shifts Change- point Detection via Derivative-Aware Gaussian Processes.arXiv preprint arXiv:2312.03176(2023). A Baseline Methods DACD.The DACD [ 35] approach is designed for CPD that in- corporates GP derivatives into the...

    2023 arXiv

  28. [2018]

    How deep are deep Gaussian processes?Journal of Machine Learning Research19, 54 (2018), 1–46

    2018

  29. [2020]

    Convergent validation of methods for the identification of psychothera- peutic phase transitions in time series of empirical and model systems.Frontiers in Psychology11 (2020), 1970

    2020

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.