arxiv: 2509.07523 · v4 · submitted 2025-09-09 · 💻 cs.LG

RoseCDL: Robust and Scalable Convolutional Dictionary Learning for Rare event and Anomaly Detection

Jad Yehya , Mansour Benbakoura , C\'edric Allain , Beno\^it Malezieux , Matthieu Kowalski , Thomas Moreau This is my paper

Pith reviewed 2026-05-18 18:12 UTC · model grok-4.3

classification 💻 cs.LG

keywords convolutional dictionary learninganomaly detectionrare eventsrobust optimizationscalable learningunsupervised detectionsignal analysis

0 comments

The pith

RoseCDL enhances convolutional dictionary learning with stochastic windowing and outlier detection to enable scalable anomaly identification in long signals.

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

Convolutional dictionary learning models local signal patterns but faces limits from computation and outlier sensitivity when signals are long. RoseCDL tackles this by training on random windows and detecting outliers during learning. The approach identifies anomalies unsupervised by measuring how well each local patch reconstructs under the learned dictionary. This matters because fields like astronomy and medicine generate signals too large for standard methods yet need to spot rare events automatically. Experiments confirm gains in both accuracy and speed on real data.

Core claim

RoseCDL is a convolutional dictionary learning algorithm that uses stochastic windowing for efficient training and inline outlier detection for robustness. It identifies anomalous and rare patterns in long signals in an unsupervised manner by relying on the local reconstruction loss, and experiments on real-world datasets demonstrate improved detection accuracy alongside reduced computational demands.

What carries the argument

The RoseCDL algorithm, which integrates stochastic windowing for training efficiency and inline outlier detection for robustness within the convolutional dictionary learning process.

If this is right

CDL can now be applied to long signals without prohibitive computation costs.
Anomalies are detected based on local reconstruction errors without needing labeled data.
The method achieves better accuracy and efficiency than standard CDL on real datasets.
Practical use becomes possible in domains requiring analysis of extensive signal data.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The same windowing and outlier handling ideas might apply to other dictionary learning variants for robustness.
Local reconstruction loss as an anomaly score could be tested in related time-series or image processing tasks.

Load-bearing premise

Stochastic windowing and inline outlier detection preserve the core modeling power of convolutional dictionary learning while making it insensitive to outliers and usable on very long signals.

What would settle it

A direct comparison experiment on a large signal dataset with known rare events, measuring whether RoseCDL's anomaly detection F1-score and training runtime outperform standard CDL; failure to improve on both would undermine the claims.

Figures

Figures reproduced from arXiv: 2509.07523 by Beno\^it Malezieux, C\'edric Allain, Jad Yehya, Mansour Benbakoura, Matthieu Kowalski, Thomas Moreau.

**Figure 2.** Figure 2: Raw signal X, reconstruction error, threshold and learned outlier mask on subject a02 (minute 56) of Physionet Apnea-ECG data set. Detection method is based on modified z-score (MAD), with α = 3.5. The method correctly identifies outliers blocks. The intuition behind our approach is that if da is sufficiently represented in x, then most of the patches of x should contain information relative to this patter… view at source ↗

**Figure 3.** Figure 3: Comparison of optimization runtime for RoseCDL, AlphaCSC, Sporco, and DeepCDL in 1D and 2D settings, highlighting the superior scalability and convergence speed of RoseCDL. The runtime plots show the evolution of test loss over time. The third subplot reports the dictionary recovery score at convergence for 2D data. As AlphaCSC does not 2D data, only results for Sporco, DeepCDL, and RoseCDL are shown. On R… view at source ↗

**Figure 4.** Figure 4: (a) Impact of the regularization on the recovery score over the epochs. (b) Atoms recovered [PITH_FULL_IMAGE:figures/full_fig_p008_4.png] view at source ↗

**Figure 5.** Figure 5: Comparison ofthe F1 score evolution of different methods over the epochs on rare event detection task on the (R, O, S, E, and Z). These images emulate text-like documents composed of words formed from the selected characters. We added the letter Z in a small proportion to introduce rare events. Experiments were conducted with a 10% contamination rate. Consequently, for the inline outlier detection meth… view at source ↗

**Figure 6.** Figure 6: Evolution of median recovery scores for RoseCDL on 2D data across 20 independent runs. [PITH_FULL_IMAGE:figures/full_fig_p009_6.png] view at source ↗

**Figure 7.** Figure 7: Learned atoms with and without outliers detection method, on 10 bad trials of subject a02 of dataset Physionet Apnea-ECG. RoseCDL on real-world data To assess the efficiency of outlier detection methods on realworld data, we utilized the Physionet Apnea-ECG dataset [43]. Notably, no preprocessing was applied to the signals, thereby minimizing manual interventions and ensuring the integrity of the raw … view at source ↗

read the original abstract

Detecting rare events and anomalies in large-scale signals is essential in fields such as astronomy, physical simulations, and biomedical science. In many cases, this problem naturally decomposes into identifying common local patterns and detecting deviations that correspond to anomalies. Convolutional Dictionary Learning (CDL) is a powerful tool for modeling local structures, but its adoption for this task has been limited by computational demands and sensitivity to outliers. We introduce RoseCDL, a novel CDL algorithm designed for robust and scalable modeling of signal pattern distribution. RoseCDL leverages stochastic windowing for efficient training and incorporates inline outlier detection to enhance robustness. This enables unsupervised identification of anomalous and rare patterns in long signals based on the local reconstruction loss. Experiments on real-world datasets show that RoseCDL delivers improved detection accuracy and computational efficiency, making CDL practical for challenging detection tasks in large-scale signal analysis.

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note private letter to a colleague

RoseCDL adds stochastic windowing and inline outlier detection to make convolutional dictionary learning scale to long signals for anomaly detection, with the full experiments apparently supporting the efficiency and accuracy claims.

read the letter

The main thing here is that RoseCDL combines stochastic windowing during training with outlier detection built into the loop. This targets the two big barriers that have kept convolutional dictionary learning from seeing much use on long recordings: compute cost and sensitivity to rare events. The result is a way to score anomalies from local reconstruction loss without having to process the entire signal at once or clean the data first.

Referee Report

0 major / 3 minor

Summary. The paper introduces RoseCDL, a novel convolutional dictionary learning (CDL) algorithm for robust and scalable modeling of local signal patterns to detect rare events and anomalies in long signals. It combines stochastic windowing for efficient training on large-scale data with inline outlier detection to reduce sensitivity to anomalies during learning, enabling unsupervised anomaly scoring via local reconstruction loss. Experiments on real-world datasets are presented to demonstrate gains in detection accuracy and computational efficiency relative to standard CDL approaches.

Significance. If the reported results hold under scrutiny, RoseCDL would meaningfully extend the practical utility of CDL to anomaly detection tasks in domains such as astronomy, physical simulations, and biomedical signal analysis by mitigating the computational scaling issues and outlier sensitivity that have historically limited its adoption. The algorithmic design, convergence analysis under stochastic sampling, and empirical validation on real data constitute a coherent contribution.

minor comments (3)

[§4.1 and Table 2] §4.1 and Table 2: the quantitative results would benefit from explicit reporting of the number of independent runs, standard deviations, and any statistical significance tests to support the claims of consistent accuracy and efficiency improvements.
[§3.3, Algorithm 1] §3.3, Algorithm 1: the pseudocode for the inline outlier detection step could include a brief note on the chosen threshold selection procedure to aid reproducibility.
[Figure 3] Figure 3: axis labels and legends are slightly undersized, reducing readability when the figure is viewed at standard print size.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for the positive assessment of RoseCDL, the recognition of its algorithmic contributions including stochastic windowing, inline outlier detection, and convergence analysis, and the recommendation for minor revision. We are pleased that the potential to extend CDL's utility to large-scale anomaly detection in astronomy, physical simulations, and biomedical signals was highlighted.

Circularity Check

0 steps flagged

No significant circularity

full rationale

The paper introduces RoseCDL as an algorithmic construction extending convolutional dictionary learning with stochastic windowing and inline outlier detection for scalability and robustness. The central claim that this enables unsupervised anomaly detection via local reconstruction loss follows directly from the method's design and is validated through convergence behavior and real-world experiments rather than any self-referential derivation. No load-bearing steps reduce by construction to fitted parameters renamed as predictions, self-citation chains, or ansatzes smuggled from prior author work; the result remains self-contained with independent empirical support.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 1 invented entities

The abstract relies on the standard premise that convolutional dictionary learning can capture local signal structure and that reconstruction loss can serve as an anomaly score; no free parameters or new physical entities are named.

axioms (1)

domain assumption Convolutional dictionary learning is a powerful tool for modeling local structures in signals
Directly stated in the abstract as the starting point for the new algorithm.

invented entities (1)

RoseCDL algorithm no independent evidence
purpose: Robust and scalable CDL variant for anomaly detection
Newly proposed method whose components are described at the level of the abstract.

pith-pipeline@v0.9.0 · 5702 in / 1314 out tokens · 35188 ms · 2026-05-18T18:12:13.216027+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

min D Ex [min Z F(D,Z;x)] with F = ½‖x−D∗Z‖² + λ‖Z‖1 and trimmed variant eF using Pβ = {τ | F < β}
IndisputableMonolith/Foundation/AbsoluteFloorClosure.lean absolute_floor_iff_bare_distinguishability unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

inline outlier detection via modified z-score / MAD on patch reconstruction errors

What do these tags mean?

matches: The paper's claim is directly supported by a theorem in the formal canon.
supports: The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends: The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses: The paper appears to rely on the theorem as machinery.
contradicts: The paper's claim conflicts with a theorem or certificate in the canon.
unclear: Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

47 extracted references · 47 canonical work pages · 1 internal anchor

[1]

ECG-based heartbeat classification for arrhythmia detection: A survey.Computer Methods and Programs in Biomedicine, 127:144–164, 2016

Luz Luz, Eduardo Jos´e da S, William Robson Schwartz, Guillermo C´amara-Ch´avez, and David Menotti. ECG-based heartbeat classification for arrhythmia detection: A survey.Computer Methods and Programs in Biomedicine, 127:144–164, 2016. 1

work page 2016
[2]

Haeffele, and Ren´e Vidal

Florence Yellin, Benjamin D. Haeffele, and Ren´e Vidal. Blood cell detection and counting in holographic lens-free imaging by convolutional sparse dictionary learning and coding. InIEEE International Symposium on Biomedical Imaging (ISBI), Melbourne, Australia, 2017. 1

work page 2017
[3]

Giavalisco and the GOODS Teams

M. Giavalisco and the GOODS Teams. The Great Observatories Origins Deep Survey: Initial Results From Optical and Near-Infrared Imaging.The Astrophysical Journal Letters, 600(2): L93, 2004. 1

work page 2004
[4]

Ra- jendra Acharya

Fatma Murat, Ozal Yildirim, Muhammed Talo, Ulas Baran Baloglu, Yakup Demir, and U. Ra- jendra Acharya. Application of deep learning techniques for heartbeats detection using ECG signals-analysis and review.Computers in Biology and Medicine, 120:103726, May 2020. 1

work page 2020
[5]

Transforming Blood Cell Detection and Classification with Advanced Deep Learning Models: A Comparative Study, October 2024

Shilpa Choudhary, Sandeep Kumar, Pammi Sri Siddhaarth, and Guntu Charitasri. Transforming Blood Cell Detection and Classification with Advanced Deep Learning Models: A Comparative Study, October 2024

work page 2024
[6]

Cornu, P

D. Cornu, P. Salom´e, B. Semelin, A. Marchal, J. Freundlich, S. Aicardi, X. Lu, G. Sainton, F. Mertens, F. Combes, and C. Tasse. YOLO-CIANNA: Galaxy detection with deep learning in radio data - I. A new YOLO-inspired source detection method applied to the SKAO SDC1. Astronomy & Astrophysics, 690:A211, October 2024. 1

work page 2024
[7]

Roger Grosse, Rajat Raina, Helen Kwong, and Andrew Y . Ng. Shift-Invariant Sparse Coding for Audio Classification.Cortex, 8:9, 2007. 1, 2

work page 2007
[8]

Multivariate convolutional sparse coding for electromagnetic brain signals.Advances in Neural Information Processing Systems, 31:3292–3302, 2018

Tom Dupr´e la Tour, Thomas Moreau, Mainak Jas, and Alexandre Gramfort. Multivariate convolutional sparse coding for electromagnetic brain signals.Advances in Neural Information Processing Systems, 31:3292–3302, 2018. 1, 4, 6

work page 2018
[9]

Convolutional dictionary learning via local processing

Vardan Papyan, Yaniv Romano, Jeremias Sulam, and Michael Elad. Convolutional dictionary learning via local processing. InProceedings of the IEEE International Conference on Computer Vision (ICCV), Oct 2017. 1, 2

work page 2017
[10]

Efficient algorithms for convolutional sparse representations.IEEE Transac- tions on Image Processing, 25(1):301–315, 2015

Brendt Wohlberg. Efficient algorithms for convolutional sparse representations.IEEE Transac- tions on Image Processing, 25(1):301–315, 2015. 2, 3

work page 2015
[11]

DiCoDiLe: Distributed Convolutional Dictionary Learning.IEEE Transactions on Pattern Analysis and Machine Intelligence, 2020

Thomas Moreau and Alexandre Gramfort. DiCoDiLe: Distributed Convolutional Dictionary Learning.IEEE Transactions on Pattern Analysis and Machine Intelligence, 2020. 2, 3, 6, 16

work page 2020
[12]

Sliding window strategy for convolutional spike sorting with lasso: Algorithm, theoretical guar- antees and complexity.Acta Appl

Laurent Dragoni, R ´emi Flamary, Karim Lounici, and Patricia Reynaud-Bouret. Sliding window strategy for convolutional spike sorting with lasso: Algorithm, theoretical guar- antees and complexity.Acta Appl. Math., 179(1), June 2022. ISSN 0167-8019. doi: 10.1007/s10440-022-00494-x. URL https://doi.org/10.1007/s10440-022-00494-x . 2

work page doi:10.1007/s10440-022-00494-x 2022
[13]

Learning fast approximations of sparse coding

Karol Gregor and Yann LeCun. Learning fast approximations of sparse coding. InProceedings of the 27th International Conference on International Conference on Machine Learning, ICML’10, page 399–406, Madison, WI, USA, 2010. Omnipress. ISBN 9781605589077. 2 11

work page 2010
[14]

Hao Tang, Hong Liu, Wei Xiao, and Nicu Sebe. When Dictionary Learning Meets Deep Learning: Deep Dictionary Learning and Coding Network for Image Recognition With Limited Data.IEEE Transactions on Neural Networks and Learning Systems, 32(5):2129–2141, May

work page
[15]

Scalable convolutional dictionary learning with constrained recurrent sparse auto-encoders

Bahareh Tolooshams, Sourav Dey, and Demba Ba. Scalable convolutional dictionary learning with constrained recurrent sparse auto-encoders. In2018 IEEE 28th International Workshop on Machine Learning for Signal Processing (MLSP), pages 1–6, 2018. doi: 10.1109/MLSP.2018. 8516996

work page doi:10.1109/mlsp.2018 2018
[16]

Bahareh Tolooshams and Demba E. Ba. Stable and interpretable unrolled dictionary learn- ing.Transactions on Machine Learning Research, 2022. ISSN 2835-8856. URL https: //openreview.net/forum?id=e3S0Bl2RO8. 2, 4, 6, 10

work page 2022
[17]

Understanding approximate and unrolled dictionary learning for pattern recovery.International Conference on Learning Representations, 2022

Benoˆıt Mal´ezieux, Thomas Moreau, and Matthieu Kowalski. Understanding approximate and unrolled dictionary learning for pattern recovery.International Conference on Learning Representations, 2022. 2, 4, 10

work page 2022
[18]

Online learning for matrix factorization and sparse coding.Journal of Machine Learning Research, 11(1), 2010

Julien Mairal, Francis Bach, Jean Ponce, and Guillermo Sapiro. Online learning for matrix factorization and sparse coding.Journal of Machine Learning Research, 11(1), 2010. 2

work page 2010
[19]

Dictionary learning for massive matrix factorization

Arthur Mensch, Julien Mairal, Bertrand Thirion, and Ga ¨el Varoquaux. Dictionary learning for massive matrix factorization. InInternational Conference on Machine Learning, pages 1737–1746. PMLR, 2016

work page 2016
[20]

Online Convolutional Dictionary Learning

Jialin Liu, Cristina Garcia-Cardona, Brendt Wohlberg, and Wotao Yin. Online Convolutional Dictionary Learning. InInternational Conference on Image Processing (ICIP), pages 1707– 1711, Beijing, China, 2017

work page 2017
[21]

Slice-Based Online Convolutional Dictionary Learning.IEEE Transactions on Cybernetics, pages 1–14, 2019

Yijie Zeng, Jichao Chen, and Guang-Bin Huang. Slice-Based Online Convolutional Dictionary Learning.IEEE Transactions on Cybernetics, pages 1–14, 2019. 2

work page 2019
[22]

Sparse and spurious: Dictionary learning with noise and outliers.IEEE Transactions on Information Theory, 61(11):6298–6319, 2015

R´emi Gribonval, Rodolphe Jenatton, and Francis Bach. Sparse and spurious: Dictionary learning with noise and outliers.IEEE Transactions on Information Theory, 61(11):6298–6319, 2015. doi: 10.1109/TIT.2015.2472522. 2, 10

work page doi:10.1109/tit.2015.2472522 2015
[23]

Learning the morphology of brain signals using alpha-stable convolutional sparse coding.Advances in Neural Information Processing Systems, pages 1099–1108, 2017

Mainak Jas, Tom Dupr ´e La Tour, Umut S ¸ims ¸ekli, and Alexandre Gramfort. Learning the morphology of brain signals using alpha-stable convolutional sparse coding.Advances in Neural Information Processing Systems, pages 1099–1108, 2017. 2

work page 2017
[24]

Kauffmann, Robert A

Lukas Ruff, Jacob R. Kauffmann, Robert A. Vandermeulen, Gregoire Montavon, Wojciech Samek, Marius Kloft, Thomas G. Dietterich, and Klaus-Robert Muller. A Unifying Review of Deep and Shallow Anomaly Detection.Proceedings of the IEEE, 109(5):756–795, May

work page
[25]

doi: 10.1109/JPROC.2021.3052449

ISSN 0018-9219, 1558-2256. doi: 10.1109/JPROC.2021.3052449. URL https: //ieeexplore.ieee.org/document/9347460/. 2

work page doi:10.1109/jproc.2021.3052449 2021
[26]

Anomaly detection in time series: a comprehensive evaluation.Proceedings of the VLDB Endowment, 15(9):1779–1797, May

Sebastian Schmidl, Phillip Wenig, and Thorsten Papenbrock. Anomaly detection in time series: a comprehensive evaluation.Proceedings of the VLDB Endowment, 15(9):1779–1797, May

work page
[27]

doi: 10.14778/3538598.3538602

ISSN 2150-8097. doi: 10.14778/3538598.3538602. URL https://dl.acm.org/doi/ 10.14778/3538598.3538602. 2

work page doi:10.14778/3538598.3538602
[28]

A comprehensive survey on rare event prediction, 2024

Chathurangi Shyalika, Ruwan Wickramarachchi, and Amit Sheth. A comprehensive survey on rare event prediction, 2024. URLhttps://arxiv.org/abs/2309.11356. 2

work page arXiv 2024
[29]

Least median of squares regression.Journal of the American statistical association, 79(388):871–880, 1984

Peter J Rousseeuw. Least median of squares regression.Journal of the American statistical association, 79(388):871–880, 1984. 2

work page 1984
[30]

John wiley & sons, 2005

Peter J Rousseeuw and Annick M Leroy.Robust regression and outlier detection. John wiley & sons, 2005. 2

work page 2005
[31]

Sparse least trimmed squares regression for analyzing high-dimensional large data sets.The Annals of Applied Statistics, pages 226–248,

Andreas Alfons, Christophe Croux, and Sarah Gelper. Sparse least trimmed squares regression for analyzing high-dimensional large data sets.The Annals of Applied Statistics, pages 226–248,

work page
[32]

Modified least trimmed quan- tile regression to overcome effects of leverage points.Mathematical Problems in Engineering, 2020:1–13, 2020

Habshah Midi, Taha Alshaybawee, and Mohammed Alguraibawi. Modified least trimmed quan- tile regression to overcome effects of leverage points.Mathematical Problems in Engineering, 2020:1–13, 2020. 2

work page 2020
[33]

Sparse least trimmed squares regression with compositional covariates for high-dimensional data.Bioinformatics, 37(21):3805–3814, 2021

Gianna Serafina Monti and Peter Filzmoser. Sparse least trimmed squares regression with compositional covariates for high-dimensional data.Bioinformatics, 37(21):3805–3814, 2021. 2

work page 2021
[34]

Robust regression shrinkage and consistent variable selection through the lad-lasso.Journal of Business & Economic Statistics, 25(3): 347–355, 2007

Hansheng Wang, Guodong Li, and Guohua Jiang. Robust regression shrinkage and consistent variable selection through the lad-lasso.Journal of Business & Economic Statistics, 25(3): 347–355, 2007. 2

work page 2007
[35]

The Trimmed Lasso: Sparsity and Robustness

Dimitris Bertsimas, Martin S Copenhaver, and Rahul Mazumder. The trimmed lasso: Sparsity and robustness.arXiv preprint arXiv:1708.04527, 2017. 2

work page internal anchor Pith review Pith/arXiv arXiv 2017
[36]

Deep k-svd denoising.IEEE Transactions on Image Processing, 30:5944–5955, 2021

Meyer Scetbon, Michael Elad, and Peyman Milanfar. Deep k-svd denoising.IEEE Transactions on Image Processing, 30:5944–5955, 2021. 3

work page 2021
[37]

Deep convolutional dictionary learning for image denoising

Hongyi Zheng, Hongwei Yong, and Lei Zhang. Deep convolutional dictionary learning for image denoising. InIEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 630–641, 2021

work page 2021
[38]

Deep M2cdl: Deep multi- scale multi-modal convolutional dictionary learning network.IEEE transactions on pattern analysis and machine intelligence, 46(5):2770–2787, 2023

Xin Deng, Jingyi Xu, Fangyuan Gao, Xiancheng Sun, and Mai Xu. Deep M2cdl: Deep multi- scale multi-modal convolutional dictionary learning network.IEEE transactions on pattern analysis and machine intelligence, 46(5):2770–2787, 2023. 3

work page 2023
[39]

A Fast Iterative Shrinkage-Thresholding Algorithm for Linear Inverse Problems.SIAM Journal on Imaging Sciences, 2(1):183–202, January 2009

Amir Beck and Marc Teboulle. A Fast Iterative Shrinkage-Thresholding Algorithm for Linear Inverse Problems.SIAM Journal on Imaging Sciences, 2(1):183–202, January 2009. ISSN 1936-4954. doi: 10.1137/080716542. URL https://epubs.siam.org/doi/10.1137/ 080716542. 3

work page doi:10.1137/080716542 2009
[40]

fast iterative shrinkage/thresholding algorithm

Antonin Chambolle and Ch Dossal. On the convergence of the iterates of the “fast iterative shrinkage/thresholding algorithm”.Journal of Optimization theory and Applications, 166: 968–982, 2015. 3

work page 2015
[41]

Painless stochastic gradient: Interpolation, line-search, and convergence rates

Sharan Vaswani, Aaron Mishkin, Issam Laradji, Mark Schmidt, Gauthier Gidel, and Simon Lacoste-Julien. Painless stochastic gradient: Interpolation, line-search, and convergence rates. Advances in neural information processing systems, 32:3732–3745, 2019. 4

work page 2019
[42]

Quality Press, 1993

Boris Iglewicz and David C Hoaglin.Volume 16: how to detect and handle outliers. Quality Press, 1993. 6

work page 1993
[43]

Automatic differentiation in PyTorch

Adam Paszke, Sam Gross, Soumith Chintala, Gregory Chanan, Edward Yang, Zachary DeVito, Zeming Lin, Alban Desmaison, Luca Antiga, and Adam Lerer. Automatic differentiation in PyTorch. 2017. 6

work page 2017
[44]

SPORCO: A Python package for standard and convolutional sparse represen- tations

Brendt Wohlberg. SPORCO: A Python package for standard and convolutional sparse represen- tations. InProceedings of the 15th Python in Science Conference, pages 1–8, Austin, TX, USA, July 2017. doi: 10.25080/shinma-7f4c6e7-001. 6

work page doi:10.25080/shinma-7f4c6e7-001 2017
[45]

The apnea-ecg database

Thomas Penzel, George B Moody, Roger G Mark, Ary L Goldberger, and J Hermann Peter. The apnea-ecg database. InComputers in Cardiology 2000. Vol. 27 (Cat. 00CH37163), pages 255–258. IEEE, 2000. 9

work page 2000
[46]

Tsb-uad: an end-to-end benchmark suite for univariate time-series anomaly detection

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. Proceedings of the VLDB Endowment, 15(8):1697–1711, 2022. 10

work page 2022
[47]

recovery score

David F Crouse. On implementing 2d rectangular assignment algorithms.IEEE Transactions on Aerospace and Electronic Systems, 52(4):1679–1696, 2016. 16 13 A Analytical study Proposition 2.1(Stability of the common pattern).Consider a population of signals X composed of two patterns da and db, with activations such that the patterns do not overlap in the sig...

work page 2016