pith. sign in

arxiv: 2306.10084 · v3 · submitted 2023-06-16 · 💻 cs.LG

Convolutional and Deep Learning based techniques for Time Series Ordinal Classification

Rafael Ayll\'on-Gavil\'an , David Guijo-Rubio , Pedro Antonio Guti\'errez , Anthony Bagnall , C\'esar Herv\'as-Mart\'inez This is my paper

Pith reviewed 2026-05-24 08:30 UTC · model grok-4.3

classification 💻 cs.LG
keywords time series classificationordinal classificationconvolutional neural networksdeep learningordinal metricsbenchmarking
0
0 comments X

The pith

Adapting convolutional and deep learning methods to respect ordinal label order improves time series classification performance.

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

The paper establishes a first benchmark for time series ordinal classification by adapting existing convolutional and deep learning TSC methods to exploit ordered class labels. It selects 29 ordinal problems from standard archives and compares the adapted ordinal versions against nominal baselines using metrics that penalize distant misclassifications. The central finding is that the ordinal adaptations yield significantly better results on those metrics, showing that ignoring label order wastes information present in many time series tasks. A sympathetic reader would care because many real classification problems involve naturally ordered categories such as severity levels or quality grades, and the work provides concrete evidence that standard TSC pipelines can be modified to use that structure.

Core claim

Ordinal adaptations of convolutional and deep learning TSC techniques achieve significantly better performance than their nominal counterparts on ordinal evaluation metrics across 29 selected time series problems drawn from two archives.

What carries the argument

Adaptations of convolutional and deep learning TSC models that incorporate ordinal label structure, typically through modified loss functions or output layers that respect the ordering of classes.

If this is right

  • Nominal TSC pipelines can be upgraded for ordinal problems by changing only the loss or output handling.
  • Ordinal metrics should become standard when evaluating classifiers on ordered time series labels.
  • Existing deep learning TSC architectures already contain the capacity to model label order once the training objective is adjusted.
  • The performance gap demonstrates that label ordering supplies usable signal beyond what the time series values alone provide.

Where Pith is reading between the lines

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

  • Similar ordinal adaptations could be tested on non-deep TSC methods such as shapelets or interval features to check whether the benefit is architecture-specific.
  • Domains like medical monitoring or industrial quality control, where time series often carry severity grades, become immediate candidates for these methods.
  • The benchmark sets a baseline that future TSOC work can use to measure progress without re-selecting datasets.

Load-bearing premise

The 29 chosen ordinal time series problems are representative of the broader class of such tasks and the method adaptations correctly exploit ordering without introducing new biases.

What would settle it

Running the same ordinal and nominal methods on a fresh collection of time series problems whose labels have a clear order and measuring whether the ordinal versions lose their advantage on ordinal metrics.

Figures

Figures reproduced from arXiv: 2306.10084 by Anthony Bagnall, C\'esar Herv\'as-Mart\'inez, David Guijo-Rubio, Pedro Antonio Guti\'errez, Rafael Ayll\'on-Gavil\'an.

Figure 1
Figure 1. Figure 1: Example of time series extracted from the [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 3
Figure 3. Figure 3: IM architecture. Two pipelines are applied in par [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: O-IN architecture, two blocks of three IMs with [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Comparison between the TSOC methodologies [PITH_FULL_IMAGE:figures/full_fig_p009_5.png] view at source ↗
Figure 7
Figure 7. Figure 7: Boxplot of relative MAEs. O-ROCKET O-MiniROCKET O-MultiROCKET O-ResNet O-InceptionTime O-LITETime O-CNN HC2 LogReg XGB TSF Methods 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 MAE [PITH_FULL_IMAGE:figures/full_fig_p010_7.png] view at source ↗
Figure 6
Figure 6. Figure 6: Run time in milliseconds (log scale average over [PITH_FULL_IMAGE:figures/full_fig_p010_6.png] view at source ↗
Figure 9
Figure 9. Figure 9: Multi-Comparison Matrix between ordinal methodologies in MAE. In each cell, three values are provided: 1) [PITH_FULL_IMAGE:figures/full_fig_p011_9.png] view at source ↗
read the original abstract

Time Series Classification (TSC) covers the supervised learning problem where input data is provided in the form of series of values observed through repeated measurements over time, and whose objective is to predict the category to which they belong. When the class values are ordinal, classifiers that take this into account can perform better than nominal classifiers. Time Series Ordinal Classification (TSOC) is the field covering this gap, yet unexplored in the literature. There are a wide range of time series problems showing an ordered label structure, and TSC techniques that ignore the order relationship discard useful information. Hence, this paper presents a first benchmarking of TSOC methodologies, exploiting the ordering of the target labels to boost the performance of current TSC state-of-the-art. Both convolutional- and deep learning-based methodologies (among the best performing alternatives for nominal TSC) are adapted for TSOC. For the experiments, a selection of 29 ordinal problems from two well-known archives has been made. In this way, this paper contributes to the establishment of the state-of-the-art in TSOC. The results obtained by ordinal versions are found to be significantly better than current nominal TSC techniques in terms of ordinal performance metrics, outlining the importance of considering the ordering of the labels when dealing with this kind of problems.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 0 minor

Summary. The paper claims to introduce Time Series Ordinal Classification (TSOC) as an unexplored extension of TSC, adapt convolutional and deep learning methods to exploit ordinal label structure, select 29 ordinal problems from two archives, and demonstrate that the ordinal adaptations yield statistically significant improvements over nominal TSC baselines on ordinal performance metrics.

Significance. If the experimental claims hold after full validation, the work would establish the first benchmark for TSOC and provide evidence that incorporating label order improves performance on ordered time series tasks. It receives credit for targeting an overlooked problem and adapting existing SOTA TSC architectures rather than proposing entirely new ones.

major comments (3)
  1. [Abstract] Abstract: the claim that ordinal versions are 'significantly better' on ordinal metrics supplies no information on which ordinal metrics were used, which statistical tests were applied, or whether any correction for multiple comparisons was performed across methods and datasets.
  2. [Problem selection] Problem selection paragraph: the statement that 'a selection of 29 ordinal problems from two well-known archives has been made' provides no selection criteria, class distribution statistics, or justification that these problems are representative of the broader TSOC task space.
  3. [Method adaptations] Method description: the adaptations of CNN/DL techniques for ordinal structure are described only at a high level; the manuscript must specify exactly how ordinal information is injected (loss function, output layer encoding, or post-processing) to support the claim that gains arise from ordinal modeling rather than hyperparameter or architectural differences.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for their constructive comments. We address each major point below and will incorporate the suggested clarifications in the revised manuscript.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the claim that ordinal versions are 'significantly better' on ordinal metrics supplies no information on which ordinal metrics were used, which statistical tests were applied, or whether any correction for multiple comparisons was performed across methods and datasets.

    Authors: We agree the abstract is insufficiently precise. The revised abstract will explicitly name the ordinal metrics (MAE and AMAE), state that statistical significance was assessed via the Wilcoxon signed-rank test, and note that no additional multiple-comparison correction was applied beyond the pairwise ordinal-vs-nominal comparisons per dataset, consistent with standard TSC benchmarking practice. revision: yes

  2. Referee: [Problem selection] Problem selection paragraph: the statement that 'a selection of 29 ordinal problems from two well-known archives has been made' provides no selection criteria, class distribution statistics, or justification that these problems are representative of the broader TSOC task space.

    Authors: The observation is correct. We will expand the relevant section to document the selection criteria (datasets from UCR/UEA archives possessing naturally ordered class labels with three or more categories), include a table of class distributions, and justify representativeness by domain coverage (medical, industrial, environmental). revision: yes

  3. Referee: [Method adaptations] Method description: the adaptations of CNN/DL techniques for ordinal structure are described only at a high level; the manuscript must specify exactly how ordinal information is injected (loss function, output layer encoding, or post-processing) to support the claim that gains arise from ordinal modeling rather than hyperparameter or architectural differences.

    Authors: We accept that greater specificity is required. The revised methods section will detail, for each technique, the exact injection mechanism (e.g., replacement of cross-entropy with an ordinal regression loss, use of a single-output threshold model instead of softmax, or post-hoc ordering enforcement) so that performance differences can be attributed to ordinal modeling. revision: yes

Circularity Check

0 steps flagged

No circularity: purely empirical benchmarking study

full rationale

The paper performs an empirical comparison by adapting convolutional and deep learning TSC methods to ordinal settings and evaluating them on 29 problems selected from existing archives. No derivations, equations, fitted parameters presented as predictions, or self-citation chains appear in the central claims. The headline result rests on experimental outcomes rather than any reduction of outputs to inputs by construction. This matches the default case of a self-contained empirical study.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Empirical benchmarking study with no mathematical derivations. Relies on standard supervised learning assumptions such as representative sampling from the chosen archives and that ordinal metrics correctly capture the desired ordering property.

pith-pipeline@v0.9.0 · 5781 in / 1016 out tokens · 55672 ms · 2026-05-24T08:30:25.141109+00:00 · methodology

discussion (0)

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

Reference graph

Works this paper leans on

85 extracted references · 85 canonical work pages · 2 internal anchors

  1. [1]

    Detecting asthma control level using feature-based time series classification,

    R. Khasha, M. M. Sepehri, and N. Taherkhani, “Detecting asthma control level using feature-based time series classification,”Applied Soft Computing, vol. 111, 2021

    work page 2021

  2. [2]

    Financial time series forecasting with deep learning : A systematic literature review: 2005-2019,

    O. B. Sezer, M. U. Gudelek, and A. M. Ozbayoglu, “Financial time series forecasting with deep learning : A systematic literature review: 2005-2019,” Applied Soft Computing, vol. 90, 2020

    work page 2005

  3. [3]

    Robust landsat-based crop time series modelling,

    D. P . Roy and L. Yan, “Robust landsat-based crop time series modelling,” Remote Sensing of Environment, vol. 238, no. SI, 2020

    work page 2020

  4. [4]

    Spontaneous termination of atrial fibrillation: A chal- lenge from physionet and computers in cardiology,

    G. Moody, “Spontaneous termination of atrial fibrillation: A chal- lenge from physionet and computers in cardiology,” in Computers in Cardiology 2004 , ser. Computing in Cardiology Series, vol. 31. IEEE, 2004, pp. 101–104

    work page 2004

  5. [5]

    Detecting forged alcohol non-invasively through vibrational spectroscopy and machine learning,

    J. Large, E. K. Kemsley, N. Wellner, I. Goodall, and A. Bag- nall, “Detecting forged alcohol non-invasively through vibrational spectroscopy and machine learning,” in 22nd Pacific-Asia Advances in Knowledge Discovery and Data Mining , 2018, pp. 298–309

    work page 2018

  6. [6]

    Nonlinear spiking neural systems with autapses for predicting chaotic time series,

    Q. Liu, H. Peng, L. Long, J. Wang, Q. Yang, M. J. P ´erez-Jim´enez, and D. Orellana-Mart ´ın, “Nonlinear spiking neural systems with autapses for predicting chaotic time series,” IEEE Transactions on Cybernetics, pp. 1–13, 2023

    work page 2023

  7. [7]

    A survey on graph neural networks for time series: Forecasting, classification, imputation, and anomaly detection,

    M. Jin, H. Y. Koh, Q. Wen, D. Zambon, C. Alippi, G. Webb, I. King, and S. Pan, “A survey on graph neural networks for time series: Forecasting, classification, imputation, and anomaly detection,” IEEE Transactions on Pattern Analysis and Machine Intelligence, 2024

    work page 2024

  8. [8]

    Time-series clustering based on the characterization of segment typologies,

    D. Guijo-Rubio, A. M. Dur ´an-Rosal, P . A. Guti ´errez, A. Tron- coso, and C. Herv ´as-Mart´ınez, “Time-series clustering based on the characterization of segment typologies,” IEEE Transactions on Cybernetics, vol. 51, no. 11, pp. 5409–5422, 2020

    work page 2020

  9. [9]

    A review and eval- uation of elastic distance functions for time series clustering,

    C. Holder, M. Middlehurst, and A. Bagnall, “A review and eval- uation of elastic distance functions for time series clustering,” Knowledge and Information Systems, vol. 66, no. 2, pp. 765–809, 2024

    work page 2024

  10. [10]

    Deep learning for time series classifi- cation and extrinsic regression: A current survey,

    N. Mohammadi Foumani, L. Miller, C. W. Tan, G. I. Webb, G. Forestier, and M. Salehi, “Deep learning for time series classifi- cation and extrinsic regression: A current survey,”ACM Computing Surveys, vol. 56, no. 9, pp. 1–45, 2024

    work page 2024

  11. [11]

    Deep learning for time series anomaly detection: A survey,

    Z. Zamanzadeh Darban, G. I. Webb, S. Pan, C. Aggarwal, and M. Salehi, “Deep learning for time series anomaly detection: A survey,” ACM Computing Surveys, vol. 57, no. 1, pp. 1–42, 2024

    work page 2024

  12. [12]

    The great time series classification bake off: a review and experimental eval- uation of recent algorithmic advances,

    A. Bagnall, J. Lines, A. Bostrom, J. Large, and E. Keogh, “The great time series classification bake off: a review and experimental eval- uation of recent algorithmic advances,”Data Mining and Knowledge Discovery, vol. 31, no. 3, pp. 606–660, 2017. IEEE TRANSACTIONS ON CYBERNETICS 12

    work page 2017

  13. [13]

    Bake off redux: a review and experimental evaluation of recent time series classifi- cation algorithms,

    M. Middlehurst, P . Sch ¨afer, and A. Bagnall, “Bake off redux: a review and experimental evaluation of recent time series classifi- cation algorithms,” Data Mining and Knowledge Discovery , vol. 38, pp. 1958–2031, 2024

    work page 1958

  14. [14]

    Fuzzy cognitive map-driven comprehensive time-series classification,

    A. Jastrzebska, G. N ´apoles, W. Homenda, and K. Vanhoof, “Fuzzy cognitive map-driven comprehensive time-series classification,” IEEE Transactions on Cybernetics, vol. 53, no. 2, pp. 1348–1359, 2023

    work page 2023

  15. [15]

    Learning a mahalanobis distance-based dynamic time warping measure for multivariate time series classification,

    J. Mei, M. Liu, Y.-F. Wang, and H. Gao, “Learning a mahalanobis distance-based dynamic time warping measure for multivariate time series classification,” IEEE Transactions on Cybernetics, vol. 46, no. 6, pp. 1363–1374, 2016

    work page 2016

  16. [16]

    Memory shapelet learning for early classification of streaming time series,

    X. Wan, L. Cen, X. Chen, Y. Xie, and W. Gui, “Memory shapelet learning for early classification of streaming time series,” IEEE Transactions on Cybernetics, vol. 54, no. 5, 2024

    work page 2024

  17. [17]

    Improving position encoding of transformers for multivariate time series classification,

    N. M. Foumani, C. W. Tan, G. I. Webb, and M. Salehi, “Improving position encoding of transformers for multivariate time series classification,” Data Mining and Knowledge Discovery, vol. 38, no. 1, pp. 22–48, 2024

    work page 2024

  18. [18]

    Reservoir computing models based on spiking neural p systems for time series classification,

    H. Peng, X. Xiong, M. Wu, J. Wang, Q. Yang, D. Orellana-Mart ´ın, and M. J. P ´erez-Jim´enez, “Reservoir computing models based on spiking neural p systems for time series classification,” Neural Networks, vol. 169, pp. 274–281, 2024

    work page 2024

  19. [19]

    Optimal online time-series segmentation,

    ´A. Carmona-Poyato, N.-L. Fern´andez-Garc´ıa, F.-J. Madrid-Cuevas, R. Mu ˜noz-Salinas, and F.-J. Romero-Ram ´ırez, “Optimal online time-series segmentation,” Knowledge and Information Systems , vol. 66, no. 4, pp. 2417–2438, 2024

    work page 2024

  20. [20]

    Motiflets: Simple and accurate detection of motifs in time series,

    P . Sch¨afer and U. Leser, “Motiflets: Simple and accurate detection of motifs in time series,” Proceedings of the VLDB Endowment , vol. 16, no. 4, pp. 725–737, 2022

    work page 2022

  21. [21]

    Discovering leitmotifs in multidimensional time series,

    ——, “Discovering leitmotifs in multidimensional time series,” arXiv preprint arXiv:2410.12293, 2024

    work page arXiv 2024

  22. [22]

    Time series classification through visual pattern recognition,

    A. Jastrzebska, “Time series classification through visual pattern recognition,” Journal of King Saud University - Computer and Infor- mation Sciences, vol. 34, no. 2, pp. 134–142, 2022

    work page 2022

  23. [23]

    Structural generative de- scriptions for time series classification,

    E. S. Garc ´ıa-Trevi˜no and J. A. Barria, “Structural generative de- scriptions for time series classification,” IEEE Transactions on Cy- bernetics, vol. 44, no. 10, pp. 1978–1991, 2014

    work page 1978

  24. [24]

    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

    work page 2019

  25. [25]

    Online ranking/collaborative filtering using the perceptron algorithm,

    E. F. Harrington, “Online ranking/collaborative filtering using the perceptron algorithm,” in Proceedings of the Twentieth International Conference on Machine Learning (ICML2003), 2003

    work page 2003

  26. [26]

    Human age estimation based on locality and ordinal information,

    C. Li, Q. Liu, W. Dong, X. Zhu, J. Liu, and H. Lu, “Human age estimation based on locality and ordinal information,” IEEE Transactions on Cybernetics, vol. 45, no. 11, pp. 2522–2534, 2015

    work page 2015

  27. [27]

    Prediction of low- visibility events due to fog using ordinal classification,

    D. Guijo-Rubio, P . A. Guti´errez, C. Casanova-Mateo, J. Sanz-Justo, S. Salcedo-Sanz, and C. Herv ´as-Mart´ınez, “Prediction of low- visibility events due to fog using ordinal classification,” Atmo- spheric Research, vol. 214, pp. 64–73, 2018

    work page 2018

  28. [28]

    Application of ordinal logistic regression analysis to identify the determinants of illness severity of covid-19 in china,

    K. Xu, M. Zhou, D. Yang, Y. Ling, K. Liu, T. Bai, Z. Cheng, and J. Li, “Application of ordinal logistic regression analysis to identify the determinants of illness severity of covid-19 in china,” Epidemiology and Infection, vol. 148, 2020

    work page 2020

  29. [29]

    Regression and ordered categorical variables,

    J. A. Anderson, “Regression and ordered categorical variables,” Journal of the Royal Statistical Society. Series B (Methodological) , vol. 46, no. 1, pp. 1–30, 1984

    work page 1984

  30. [30]

    Predictive Modelling of Bone Age through Classification and Regression of Bone Shapes

    A. Bagnall and L. Davis, “Predictive modelling of bone age through classification and regression of bone shapes,” arXiv preprint arXiv:1406.4781, 2014

    work page internal anchor Pith review Pith/arXiv arXiv 2014

  31. [31]

    Aug- mented bilinear network for incremental multi-stock time-series classification,

    M. Shabani, D. T. Tran, J. Kanniainen, and A. Iosifidis, “Aug- mented bilinear network for incremental multi-stock time-series classification,” Pattern Recognition, vol. 141, p. 109604, 2023

    work page 2023

  32. [32]

    Rocket: exceptionally fast and accurate time series classification using random convolu- tional kernels,

    A. Dempster, F. Petitjean, and G. I. Webb, “Rocket: exceptionally fast and accurate time series classification using random convolu- tional kernels,” Data Mining and Knowledge Discovery , vol. 34, pp. 1454–1495, 2020

    work page 2020

  33. [33]

    A very fast (almost) deterministic transform for time series classification angus dempster, daniel f. schmidt, geof- frey i. webb,

    A. Dempster, “A very fast (almost) deterministic transform for time series classification angus dempster, daniel f. schmidt, geof- frey i. webb,” arXiv preprint arXiv:2012.08791, 2020

    work page arXiv 2012

  34. [34]

    Multi- rocket: multiple pooling operators and transformations for fast and effective time series classification,

    C. W. Tan, A. Dempster, C. Bergmeir, and G. I. Webb, “Multi- rocket: multiple pooling operators and transformations for fast and effective time series classification,” Data Mining and Knowledge Discovery, pp. 1–24, 2022

    work page 2022

  35. [35]

    Hydra: Competing convolutional kernels for fast and accurate time series classifica- tion,

    A. Dempster, D. F. Schmidt, and G. I. Webb, “Hydra: Competing convolutional kernels for fast and accurate time series classifica- tion,” Data Mining and Knowledge Discovery, vol. 37, no. 5, pp. 1779– 1805, 2023

    work page 2023

  36. [36]

    Detach- rocket: Sequential feature selection for time series classification with random convolutional kernels,

    G. Uribarri, F. Barone, A. Ansuini, and E. Frans ´en, “Detach- rocket: Sequential feature selection for time series classification with random convolutional kernels,” Data Mining and Knowledge Discovery, pp. 1–26, 2024

    work page 2024

  37. [37]

    Time series classification from scratch with deep neural networks: A strong baseline,

    Z. Wang, W. Yan, and T. Oates, “Time series classification from scratch with deep neural networks: A strong baseline,” in 2017 International joint conference on neural networks (IJCNN). IEEE, 2017, pp. 1578–1585

    work page 2017

  38. [38]

    InceptionTime: Finding AlexNet for time series classification,

    H. I. 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, pp. 1936–1962, 2020

    work page 1936

  39. [39]

    Deep learning for time series classification and extrinsic regression: A current survey,

    N. M. Foumani, L. Miller, C. W. Tan, G. I. Webb, G. Forestier, and M. Salehi, “Deep learning for time series classification and extrinsic regression: A current survey,” 2023

    work page 2023

  40. [40]

    Deep learning for time series classification using new hand-crafted convolution filters,

    A. Ismail-Fawaz, M. Devanne, J. Weber, and G. Forestier, “Deep learning for time series classification using new hand-crafted convolution filters,” in 2022 IEEE International Conference on Big Data, 2022, pp. 972–981

    work page 2022

  41. [41]

    Lite: Light inception with boosting techniques for time series classification,

    A. Ismail-Fawaz, M. Devanne, S. Berretti, J. Weber, and G. Forestier, “Lite: Light inception with boosting techniques for time series classification,” in2023 IEEE 10th International Conference on Data Science and Advanced Analytics . IEEE, 2023, pp. 1–10

    work page 2023

  42. [42]

    A review of unsuper- vised feature learning and deep learning for time-series model- ing,

    M. L ¨angkvist, L. Karlsson, and A. Loutfi, “A review of unsuper- vised feature learning and deep learning for time-series model- ing,” Pattern Recognition Letters, vol. 42, pp. 11–24, 2014

    work page 2014

  43. [43]

    Difference-guided representation learning network for multivariate time-series clas- sification,

    Q. Ma, Z. Chen, S. Tian, and W. W. Ng, “Difference-guided representation learning network for multivariate time-series clas- sification,” IEEE Transactions on Cybernetics, vol. 52, no. 6, pp. 4717– 4727, 2022

    work page 2022

  44. [44]

    Autotransformer: Automatic transformer architecture design for time series classification,

    Y. Ren, L. Li, X. Yang, and J. Zhou, “Autotransformer: Automatic transformer architecture design for time series classification,” in Pacific-Asia Conference on Knowledge Discovery and Data Mining . Springer, 2022, pp. 143–155

    work page 2022

  45. [45]

    Haqjsk: Hierarchical-aligned quantum jensen-shannon kernels for graph classification,

    L. Bai, L. Cui, Y. Wang, M. Li, J. Li, P . S. Yu, and E. R. Hancock, “Haqjsk: Hierarchical-aligned quantum jensen-shannon kernels for graph classification,” IEEE Transactions on Knowledge and Data Engineering, vol. 36, no. 11, pp. 6370–6384, 2024

    work page 2024

  46. [46]

    Guest editorial: deep neural networks for graphs: theory, models, algorithms, and applications,

    M. Li, A. Micheli, Y. G. Wang, S. Pan, P . Li ´o, G. S. Gnecco, and M. Sanguineti, “Guest editorial: deep neural networks for graphs: theory, models, algorithms, and applications,”IEEE Transactions on Neural Networks and Learning Systems, vol. 35, no. 4, pp. 4367–4372, 2024

    work page 2024

  47. [47]

    Support vector learning for ordinal regression,

    R. Herbrich, T. Graepel, and K. Obermayer, “Support vector learning for ordinal regression,” in Ninth International Conference on Artificial Neural Networks (ICANN99). IEE, 1999, pp. 97–102

    work page 1999

  48. [48]

    Support vector ordinal regression,

    W. Chu and S. Keerthi, “Support vector ordinal regression,” Neural computation, vol. 19, pp. 792–815, 2007

    work page 2007

  49. [49]

    Kernel principal com- ponent analysis: Applications, implementation and comparison,

    D. Olsson, P . Georgiev, and P . M. Pardalos, “Kernel principal com- ponent analysis: Applications, implementation and comparison,” in Models, Algorithms, and Technologies for Network Analysis, vol. 59, 2013, pp. 127–148

    work page 2013

  50. [50]

    Gaussian processes for ordinal regression,

    W. Chu and Z. Ghahramani, “Gaussian processes for ordinal regression,” Journal of Machine Learning Research , vol. 6, pp. 1019– 1041, 2005

    work page 2005

  51. [51]

    Validation based sparse gaussian processes for ordinal regression,

    P . K. Srijith, S. Shevade, and S. Sundararajan, “Validation based sparse gaussian processes for ordinal regression,” in Neural Infor- mation Processing, ICONIP 2012, vol. 7664. IEEE, 2012, pp. 409–416

    work page 2012

  52. [52]

    A probabilistic least squares approach to ordinal regression,

    P . Srijith, S. Shevade, and S. Sundararajan, “A probabilistic least squares approach to ordinal regression,” in Advances in AI , 2012, pp. 683–694

    work page 2012

  53. [53]

    Semi-supervised gaussian process ordinal regression,

    P . K. Srijith, S. Shevade, and S. Sundararajan, “Semi-supervised gaussian process ordinal regression,” in Machine Learning and Knowledge Discovery in Databases . Berlin, Heidelberg: Springer Berlin Heidelberg, 2013, pp. 144–159

    work page 2013

  54. [54]

    Orfeo: Ordinal classifier and regres- sor fusion for estimating an ordinal categorical target,

    A. M. G ´omez-Orellana, D. Guijo-Rubio, P . A. Guti´errez, C. Herv´as- Mart´ınez, and V . M. Vargas, “Orfeo: Ordinal classifier and regres- sor fusion for estimating an ordinal categorical target,”Engineering Applications of Artificial Intelligence, vol. 133, p. 108462, 2024

    work page 2024

  55. [55]

    A decision tree-based method for ordinal classification problems,

    M. Marudi, I. Ben-Gal, and G. Singer, “A decision tree-based method for ordinal classification problems,” IISE Transactions , vol. 56, no. 9, pp. 960–974, 2024

    work page 2024

  56. [56]

    Cumula- tive link models for deep ordinal classification,

    V . M. Vargas, P . A. Gutierrez, and C. Hervas-Martinez, “Cumula- tive link models for deep ordinal classification,” Neurocomputing, vol. 401, pp. 48–58, 2020. IEEE TRANSACTIONS ON CYBERNETICS 13

    work page 2020

  57. [57]

    Deep ordinal classification based on the proportional odds model,

    V . M. Vargas, P . A. Guti ´errez, and C. Herv ´as-Mart´ınez, “Deep ordinal classification based on the proportional odds model,” in From Bioinspired Systems and Biomedical Applications to Machine Learning, vol. 11487, 2019, pp. 441–451

    work page 2019

  58. [58]

    A novel deep ordinal classification approach for aesthetic quality control classification,

    R. Rosati, L. Romeo, V . M. Vargas, P . A. Guti ´errez, C. Herv ´as- Mart´ınez, and E. Frontoni, “A novel deep ordinal classification approach for aesthetic quality control classification,” Neural Com- puting and Applications, 2022

    work page 2022

  59. [59]

    Learning ordinal–hierarchical constraints for deep learning classifiers,

    R. Rosati, L. Romeo, V . M. Vargas, P . A. Gutierrez, E. Frontoni, and C. Hervas-Martinez, “Learning ordinal–hierarchical constraints for deep learning classifiers,” IEEE Transactions on Neural Networks and Learning Systems, 2024

    work page 2024

  60. [60]

    Generalised triangular distributions for ordinal deep learning: Novel proposal and optimisation,

    V . M. Vargas, A. M. Dur ´an-Rosal, D. Guijo-Rubio, P . A. Guti´errez, and C. Herv ´as-Mart´ınez, “Generalised triangular distributions for ordinal deep learning: Novel proposal and optimisation,” Informa- tion Sciences, vol. 648, p. 119606, 2023

    work page 2023

  61. [61]

    Time series ordinal classification via shapelets,

    D. Guijo-Rubio, P . A. Gutierrez, A. Bagnall, and C. Hervas- Martinez, “Time series ordinal classification via shapelets,” in2020 International Joint Conference on Neural Networks . IEEE, 2020

    work page 2020

  62. [62]

    Ordinal versus nominal time series classification,

    D. Guijo-Rubio, P . A. Guti ´errez, A. Bagnall, and C. Herv ´as- Mart´ınez, “Ordinal versus nominal time series classification,” in International Workshop on Advanced Analytics and Learning on Temporal Data. Springer, 2020, pp. 19–29

    work page 2020

  63. [63]

    Studying the effect of different lp norms in the context of time series ordinal classification,

    D. Guijo-Rubio, V . M. Vargas, P . A. Guti ´errez, and C. Herv ´as- Mart´ınez, “Studying the effect of different lp norms in the context of time series ordinal classification,” in Conference of the Spanish Association for Artificial Intelligence. Springer, 2021, pp. 44–53

    work page 2021

  64. [64]

    Regression models for ordinal data,

    P . McCullagh, “Regression models for ordinal data,” Journal of the Royal Statistical Society: Series B (Methodological) , vol. 42, no. 2, pp. 109–127, 1980

    work page 1980

  65. [65]

    Time series extrinsic regression: Predicting numeric values from time series data,

    C. W. Tan, C. Bergmeir, F. Petitjean, and G. I. Webb, “Time series extrinsic regression: Predicting numeric values from time series data,” Data Mining and Knowledge Discovery , vol. 35, no. 3, pp. 1032–1060, 2021

    work page 2021

  66. [66]

    Unsupervised feature based algorithms for time series extrinsic regression,

    D. Guijo-Rubio, M. Middlehurst, G. Arcencio, D. F. Silva, and A. Bagnall, “Unsupervised feature based algorithms for time series extrinsic regression,” Data Mining and Knowledge Discovery , pp. 1– 45, 2024

    work page 2024

  67. [67]

    On the consistency of ordinal regression methods,

    F. Pedregosa, F. Bach, and A. Gramfort, “On the consistency of ordinal regression methods,” The Journal of Machine Learning Research, vol. 18, no. 1, pp. 1769–1803, 2017

    work page 2017

  68. [68]

    Ordinal regression methods: Survey and experimental study,

    P . Guti ´errez, M. P ´erez-Ort´ız, J. S´anchez-Monedero, F. Fern ´andez- Navarro, and C. Herv ´as-Mart´ınez, “Ordinal regression methods: Survey and experimental study,” IEEE Transactions on Knowledge and Data Engineering, vol. 28, no. 1, pp. 127–146, 2016

    work page 2016

  69. [69]

    Weighted kappa loss function for multi-class classification of ordinal data in deep learning,

    J. de La Torre, D. Puig, and A. Valls, “Weighted kappa loss function for multi-class classification of ordinal data in deep learning,” Pattern Recognition Letters, vol. 105, pp. 144–154, 2018

    work page 2018

  70. [70]

    An Empirical Evaluation of Generic Convolutional and Recurrent Networks for Sequence Modeling

    S. Bai, J. Z. Kolter, and V . Koltun, “An empirical evaluation of generic convolutional and recurrent networks for sequence modeling,” arXiv preprint arXiv:1803.01271, 2018

    work page internal anchor Pith review Pith/arXiv arXiv 2018

  71. [71]

    Convolutional neural networks for time series classification,

    B. Zhao, H. Lu, S. Chen, J. Liu, and D. Wu, “Convolutional neural networks for time series classification,” Journal of Systems Engineering and Electronics, vol. 28, no. 1, pp. 162–169, 2017

    work page 2017

  72. [72]

    Deep residual learning for image recognition,

    K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in Proceedings of the IEEE conference on computer vision and pattern recognition, 2016, pp. 770–778

    work page 2016

  73. [73]

    Deep neural network ensembles for time series classification,

    H. I. Fawaz, G. Forestier, J. Weber, L. Idoumghar, and P .-A. Muller, “Deep neural network ensembles for time series classification,” in 2019 International Joint Conference on Neural Networks (IJCNN) . IEEE, 2019, pp. 1–6

    work page 2019

  74. [74]

    aeon: a python toolkit for learning from time series,

    M. Middlehurst, A. Ismail-Fawaz, A. Guillaume, C. Holder, D. Guijo-Rubio, G. Bulatova, L. Tsaprounis, L. Mentel, M. Walter, P . Sch¨afer et al. , “aeon: a python toolkit for learning from time series,” Journal of Machine Learning Research , vol. 25, no. 289, pp. 1–10, 2024

    work page 2024

  75. [75]

    The regression analysis of binary sequences,

    D. R. Cox, “The regression analysis of binary sequences,” Journal of the Royal Statistical Society: Series B (Methodological), vol. 20, no. 2, pp. 215–232, 1958

    work page 1958

  76. [76]

    Xgboost: extreme gradient boosting,

    T. Chen, T. He, M. Benesty, V . Khotilovich, Y. Tang, H. Cho, K. Chen et al. , “Xgboost: extreme gradient boosting,” R package version 0.4-2, vol. 1, no. 4, pp. 1–4, 2015

    work page 2015

  77. [77]

    A time series forest for classification and feature extraction,

    H. Deng, G. Runger, E. Tuv, and M. Vladimir, “A time series forest for classification and feature extraction,” Information Sciences, vol. 239, pp. 142–153, 2013

    work page 2013

  78. [78]

    HIVE-COTE 2.0: a new meta ensemble for time series classification,

    M. Middlehurst, J. Large, M. Flynn, J. Lines, A. Bostrom, and A. Bagnall, “HIVE-COTE 2.0: a new meta ensemble for time series classification,” Machine Learning, vol. 110, pp. 3211–3243, 2021

    work page 2021

  79. [79]

    Simultaneous short-term significant wave height and energy flux prediction using zonal multi-task evolutionary artificial neural networks,

    A. G ´omez, D. Guijo-Rubio, P . Guti´errez, and C. Herv ´as-Mart´ınez, “Simultaneous short-term significant wave height and energy flux prediction using zonal multi-task evolutionary artificial neural networks,” Renewable Energy, vol. 184, pp. 975–989, 2022

    work page 2022

  80. [80]

    A cascaded convolutional neural network for age estimation of unconstrained faces,

    J.-C. Chen, A. Kumar, R. Ranjan, V . M. Patel, A. Alavi, and R. Chellappa, “A cascaded convolutional neural network for age estimation of unconstrained faces,” in 8th International conference on biometrics theory, applications and systems. IEEE, 2016, pp. 1–8

    work page 2016

Showing first 80 references.