Pith. sign in

REVIEW 5 major objections 4 minor 43 references

Copula-based mixture model identification for subgroup clustering with imaging applications

T0 review · 5 major / 4 minor · reviewed 2026-08-08 · deepseek-v4-flash

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

Pith's one-line read An unsupervised algorithm, CBMM-GICE, identifies copula-based mixture models with cluster-specific marginals and copulas and generally outperforms Gaussian-mixture EM on synthetic, MNIST, and cardiac MRI data.

desk verdict A plausible, clearly written adaptation of GICE to copula-based mixture identification with useful experiments, but the headline claim overreaches: convergence is unproven, a key baseline is missing from the synthetic section, and an ambiguity in Algorithm 1 needs fixing before it can be trusted. read the letter →

arxiv 2502.08549 v1 pith:K2OZ6462 submitted 2025-02-12 cs.CV cs.LG

Fei Zheng , Nicolas Duchateau This is my paper · ORCID
classification cs.CVcs.LG
keywords copula-basedmixturemodelsGICEalgorithmmodelselectionfinitemixturesclusteringmedicalimagingunsupervisedlearningMNIST
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper proposes an unsupervised algorithm, CBMM-GICE, that builds a copula-based mixture model for a cohort of samples and clusters the samples into subgroups at the same time. Unlike a Gaussian mixture model, each cluster is allowed to have its own marginal distribution per dimension and its own dependence structure (copula), so skewed, heavy-tailed, or otherwise non-elliptical subgroups can be represented without artificially inflating the number of components. The authors argue that this flexibility matters in medical imaging, where patient subgroups often do not follow the same distribution family. They report that CBMM-GICE generally outperforms GMM-EM on synthetic two-cluster data, on the full MNIST database, and on cardiac magnetic resonance data for infarct territory and pattern clustering.

What carries the argument

The central object is the copula-based mixture model built on Sklar's theorem, which writes each component density as a copula density acting on the marginal CDFs times the product of the marginal densities: $p_k(x) = c_k(F_{k,1}(x_1),\dots,F_{k,D}(x_D)) \prod_{d=1}^{D} f_{k,d}(x_d)$. The machinery that carries the argument is the GICE loop: simulate the hidden labels $T$ times from the current posteriors, fuse the $T$ subgroups for each component, then apply Kolmogorov-distance decision rules to pick one marginal form per dimension from a dictionary $\mathcal{H}$ and one copula form from a dictionary $\mathcal{G}$, estimating parameters with maximum likelihood for marginals and pseudo-maximum likelihood for copulas. This form-selection step replaces the combinatorial search over all combinations of candidate forms. The realization time $T$ smooths the stochastic convergence at the price of runtime.

What would settle it

Run CBMM-GICE on the synthetic two-cluster CBMM ground truth of Section 4.1 (N=2000, known parameters) from many random seeds, with both K-Means and GMM initialization and T=1 versus T=10, and compare the selected marginal and copula forms and label error ratios at iteration 100; if the final forms or error ratios vary materially across seeds on the same fixed data, the method's reported advantage over GMM-EM would not transfer reliably to new cohorts.

Watch

Extended reading notes

Core claim

The paper's central claim is that the Generalized Iterative Conditional Estimation (GICE) algorithm, originally designed for switching hidden Markov models, can be adapted to identify Copula-Based Mixture Models (CBMMs) in an unsupervised way. At each iteration the algorithm simulates hidden cluster labels from the current model, groups the samples accordingly, and then selects — for each cluster, each dimension, and the cluster's dependency structure — the best marginal and copula forms from candidate dictionaries, re-estimating their parameters as it goes. Because the candidate list includes Gaussian marginals and the Gaussian copula, GMMs are a special case of the search space. The authors report that on the full MNIST database CBMM-GICE reaches an average clustering accuracy of 0.848 versus 0.824 for GMM-EM, and on cardiac infarct data it produces lower Kolmogorov distances and higher silhouette scores, supporting the conclusion that heterogeneous, non-elliptical components fit real cohorts better.

Load-bearing premise

The load-bearing premise is that the algorithm's random label-simulation loop settles on a stable, useful mixture model in practice; the paper gives no theoretical convergence proof and only monitors a goodness-of-fit index over at most 100 iterations, so different random seeds or starting points could in principle lead to different models.

Editorial extensions

If this is right

  • Since Gaussian marginals and Gaussian copulas are inside the candidate dictionaries, GMM is a special case of the model family CBMM-GICE searches, so the method can only gain flexibility relative to GMM-EM on the same data.
  • On the full MNIST database the method reports an average clustering accuracy of 0.848 against 0.824 for GMM-EM, with a best run of 0.967, indicating that the added distributional flexibility helps on overlapping clusters.
  • On cardiac infarct data the method reports lower Kolmogorov distances than GMM-EM and MMST-EM and the highest mean silhouette score (0.513 versus 0.452), which matters for subgroup discovery in clinical cohorts.
  • The realization time T provides a practical trade-off: larger T gives smoother convergence and stabler estimates, but runtime grows from about 207 s at T=1 to 1548 s at T=10 in the synthetic test.

Reading between the lines

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

  • Editorial inference: because convergence is only monitored and not proven, a natural stress test is to run many random seeds on the Section 4.1 synthetic ground truth and measure how often the selected marginal/copula forms and the label error ratio settle on the same answer.
  • Editorial inference: the real-data experiments all work in a 2D UMAP projection, so the reported accuracy reflects the projection quality as well as the mixture identification; comparing GMM-EM on the same coordinates, as the paper does, is the right way to isolate the mixture-model contribution.
  • Editorial inference: if the form-selection loop proves stable, the same scheme could extend to higher-dimensional latent spaces using vine or neural copulas, the direction the authors flag as the main obstacle to going beyond 2D.
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

5 major / 4 minor

Summary. The paper proposes CBMM-GICE, an adaptation of the Generalized Iterative Conditional Estimation (GICE) algorithm for unsupervised identification of copula-based mixture models (CBMMs) with heterogeneous marginal and copula forms. The algorithm iteratively simulates latent labels, selects marginal and copula forms from finite dictionaries, estimates parameters, and monitors convergence through a Kolmogorov distance index. The method is evaluated on synthetic two-cluster data, the full MNIST database after UMAP projection, and cardiac MRI infarct-pattern data, where it is compared with GMM-EM and, in one experiment, MMST-EM. The authors conclude that CBMM-GICE generally outperforms classic EM-identified mixtures and can identify subgroups with non-Gaussian shapes.

Significance. If the empirical claims are supported, the contribution is practically valuable: it provides a flexible parametric clustering method that performs marginal and copula form selection inside an iterative estimation scheme, avoiding exhaustive model-combination searches. The paper tests the method on a large public dataset and a clinical cohort, and its discussion of limitations is candid. The current evidence, however, does not yet fully support the breadth of the claims: there is no convergence guarantee, the pseudocode is internally inconsistent at one point, and several experimental comparisons lack uncertainty quantification and relevant baselines. The contribution is therefore promising but conditional on addressing these issues.

major comments (5)
  1. [Section 3.2 and Algorithm 1] The central claim that CBMM-GICE reliably identifies a useful CBMM presupposes that the stochastic iterations settle at a stable model. The paper states in Section 3.2 that 'It is difficult to study theoretically the convergence of GICE' and only monitors convergence via the Kolmogorov distance in Eq. (15), with a fixed iterMax=100 and stochastic label simulation with T=10 or T=50. Because no monotone objective is optimized, the sequence could in principle cycle or depend on the random seed, UMAP projection, and initialization. Table 3 reports min-max accuracy ranges that overlap substantially between GMM-EM (0.690-0.956) and CBMM-GICE (0.716-0.967), and Table 4 reports single values without repeated runs. Please provide stability evidence: repeated seeds and initializations with full distributions, sensitivity to T and iterMax, and a check for fixed points or cycles.
  2. [Algorithm 1, lines 7-8] The pseudocode in Algorithm 1 assigns alpha_k = MarginEstimation and theta_k = CopulaEstimation, which is the reverse of the definitions in Section 3 (steps 2(c)-2(d)) and of Algorithms 2 and 3, where MarginEstimation returns theta_k and CopulaEstimation returns alpha_k. This ambiguity affects the exact procedure being evaluated and must be resolved. Please correct the pseudocode or the text and state explicitly which quantity is selected by the decision rules Delta_1 and Delta_2 at each step.
  3. [Section 4.1 and Section 5] The conclusion states that CBMM-GICE was tested against GMM-EM 'both on synthetic and real medical imaging data', but the synthetic experiments in Section 4.1 compare only GICE variants with different T and initialization and include no GMM-EM baseline. The claimed general outperformance over GMM-EM is therefore not directly supported on synthetic data. Please add a GMM-EM comparison to the synthetic experiments or temper the conclusion accordingly.
  4. [Sections 4.3.1-4.3.2, Table 4 and Figure 5] The cardiac evaluation is entirely in-sample: the UMAP projection is selected by trustworthiness on the same cohort, K=3 is selected by BIC on the same LAD data, and both Kolmogorov distance and silhouette scores are computed on the data used for fitting. Table 4 reports a single run without error bars or repeated UMAP projections/seeds. The reported improvements (silhouette 0.513 vs 0.452; Kolmogorov distance 0.059 vs 0.066/0.067) therefore lack uncertainty quantification and may reflect in-sample fit. Please add repeated projections and seeds and report confidence intervals or significance tests.
  5. [Section 1 and references [13,37]] The paper identifies Online EMDL and VCMM as closely related CBMM identification methods but never compares with them in the experiments. Since these methods directly address the same marginal and copula form-selection problem, a comparison on at least the synthetic data, or an explicit statement of unavailability, would substantially strengthen the claim that the proposed identification strategy is original and advantageous.
minor comments (4)
  1. [Table 3] The header 'GMM-EN' should read 'GMM-EM'.
  2. [Section 4.1.1] The text reports running time on a '2.7MHz CPU'; this should presumably be '2.7 GHz CPU'.
  3. [Figure 5 caption] The caption says 'clusters found be GMM-EM and CBMM-GICE'; it should read 'found by GMM-EM and CBMM-GICE'.
  4. [Table A2] The parameter range for Arch12 and Arch14 is written as alpha in [1,+infinity]; if the endpoint is not attained, use [1,infinity) for clarity.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: CBMM-GICE is an empirical fitting and clustering method whose main claims are checked against external ground truth, and the cited GICE lineage is background rather than a load-bearing self-referential argument.

full rationale

The derivation chain is self-contained. The model is defined through Sklar's theorem (Eqs. 3-6), and the identification algorithm is specified as iterative conditional estimation with simulated labels (Algorithm 1), parameter estimators (Eqs. 11-12), and Kolmogorov-distance-based decision rules (Eqs. 13-14). No equation defines the output in terms of the fitted parameters or vice versa: the estimated CBMM is the output, and the evaluation is not a renamed version of the fitting objective. Clustering performance is measured against external ground truth where available: synthetic labels in Section 4.1, MNIST digit labels in Section 4.2, and coronary territory labels in Section 4.3. The Kolmogorov distance and silhouette scores are in-sample fit or cluster-quality measures and are presented as such, not as predictions. The algorithm's lineage is honestly attributed to prior GICE work [42, 8] and to ICE [33]; this is normal background citation. Moreover, the paper explicitly disclaims a theoretical convergence guarantee in Section 3.2 ('It is difficult to study theoretically the convergence of GICE') and in Section 5 ('The convergence of GICE is hard to prove theoretically'), so it does not import a convergence or uniqueness theorem from prior work as a load-bearing proof. The limitations that do exist are not circularity: Algorithm 1 lines 7-8 swap the marginal and copula estimation calls relative to Algorithms 2/3 and the text, creating an implementation ambiguity; the synthetic experiments in Section 4.1 compare GICE variants without a GMM-EM baseline, so the conclusion's phrase 'tested both on synthetic and real medical imaging data' overstates the comparison; and the MNIST accuracy ranges overlap substantially (GMM-EM 0.690-0.956 vs CBMM-GICE 0.716-0.967). These are correctness, reproducibility, and overclaiming concerns, not instances of the derivation reducing to its own inputs. The paper therefore receives a circularity score of 0.

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

The central result rests on the copula decomposition (Sklar's theorem), on the assumed existence of estimators and decision rules for each candidate form, on the choice of finite candidate dictionaries, and on an unproven convergence of the stochastic GICE iterations. The UMAP projection to 2D is also a domain assumption for the real-data evaluations.

free parameters (6)
  • Realization time T = 1, 10, 50 in experiments
    Number of simulated label sets per iteration; larger T smooths convergence at higher computational cost and affects final estimates. The paper recommends T around 10 informally.
  • UMAP KNN = 30 for MNIST, 6 for cardiac data
    Neighborhood size for the UMAP projection. It controls local versus global structure and is chosen via documentation, trustworthiness scoring, or observed best clustering results.
  • Number of clusters K = 2 synthetic, 10 MNIST, 3 LAD via BIC
    Cluster count is fixed by experimental design or selected by BIC on the projected data; the clustering output and identified forms depend on this choice.
  • Maximum iterations iterMax = 100
    Cap on GICE iterations. Convergence is not proven, so results can depend on this stopping limit.
  • Candidate form dictionaries H and G = 7 marginal forms, 7 copula forms
    The identified 'best' forms are relative to this finite dictionary chosen by the authors; forms outside the list cannot be selected.
  • Initialization strategy = K-Means or GMM-EM
    Initial parameters affect early GICE iterations. Synthetic experiments suggest convergence to similar final models, but this is not guaranteed.
assumptions (6)
  • standard math Sklar's theorem for copula decomposition of joint distributions
    Used in Equations (3) through (6) to write each component density as a copula density times the product of marginal densities; this is the theoretical foundation of the CBMM.
  • domain assumption Existence of estimators for all candidate marginal and copula forms
    Assumption 1 in Section 3 states that estimators exist for each candidate form. If any candidate form is poorly estimated, the form-selection step will be unreliable.
  • domain assumption Existence of reliable decision rules Delta_1 and Delta_2
    Assumption 2 in Section 3 assumes decision rules select the best-fit marginal and copula forms. The paper uses the Kolmogorov distance, whose model-selection consistency is not established.
  • ad hoc to paper The finite candidate dictionaries contain adequate distribution forms
    Experiments restrict H to 7 marginals and G to 7 copulas (Section 4). If the true generating forms are outside these lists, identification is only best-fit within the chosen dictionary.
  • ad hoc to paper UMAP 2D projection preserves the cluster structure needed for clustering
    All real-data experiments cluster on a 2D UMAP projection (Sections 4.2 and 4.3). The method itself is only implemented for 2D, so projection quality is load-bearing.
  • ad hoc to paper GICE converges to a stable fixed point within iterMax iterations
    Section 3.2 explicitly says theoretical convergence is difficult and only proposes monitoring via the Kolmogorov distance. The usefulness of the algorithm assumes the stochastic iterations stabilize in practice.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Copula-based mixture model identification for subgroup clustering with imaging applications." pith.science (2026). https://pith.science/paper/K2OZ6462

@misc{pith2026250208549,
  author       = {Pith},
  title        = {Pith review of: Copula-based mixture model identification for subgroup clustering with imaging applications},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/K2OZ6462}},
  note         = {Machine review of arXiv:2502.08549}
}
read the original abstract

Model-based clustering techniques have been widely applied to various application areas, while most studies focus on canonical mixtures with unique component distribution form. However, this strict assumption is often hard to satisfy. In this paper, we consider the more flexible Copula-Based Mixture Models (CBMMs) for clustering, which allow heterogeneous component distributions composed by flexible choices of marginal and copula forms. More specifically, we propose an adaptation of the Generalized Iterative Conditional Estimation (GICE) algorithm to identify the CBMMs in an unsupervised manner, where the marginal and copula forms and their parameters are estimated iteratively. GICE is adapted from its original version developed for switching Markov model identification with the choice of realization time. Our CBMM-GICE clustering method is then tested on synthetic two-cluster data (N=2000 samples) with discussion of the factors impacting its convergence. Finally, it is compared to the Expectation Maximization identified mixture models with unique component form on the entire MNIST database (N=70000), and on real cardiac magnetic resonance data (N=276) to illustrate its value for imaging applications.

Figures

Figures reproduced from arXiv: 2502.08549 by the authors.

Figure 1
Figure 1. Synthetic experiment (N=2000 samples) to evaluate the performance of GICE on non-Gaussian CBMM identification. better-fit initialization than K-Means, by comparing the Kolmogorov distance and error ratio of the two GICE settings “T = 10, init: K-Means” and “T = 10, init: GMM” at the start of the algorithm in Figure B1. However, after 30 itera￾tions they converged to the same level and give very similar estimated CBM… view at source ↗
Figure 2
Figure 2. Synthetic experiment (N=2000 samples) to evaluate the performance of GICE on GMM identification. 4.1.2. Test on simulated data of GMM This series of tests aims to check if GICE also works for the identification of GMM, which can be considered a special case of CBMM. 2000 samples were simulated from a pre-defined two-component CBMM with Gaussian marginals and copulas (thus, it is actually a GMM), see Figure 2a for th… view at source ↗
Figure 3
Figure 3. Experiment on the MNIST dataset (N=70000 samples, 10 clusters, 2D projection obtained by UMAP with KNN=30), to evaluate the performance of GICE on more than two clusters. Each point corresponds to the image of a digit, colored by its corresponding ground truth label. goodness of fit. The GMM-EM performance is quite close to the one (0.825) reported in [28] [PITH_FULL_IMAGE:figures/full_fig_p018_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Clustering of the projected infarct segments in LAD and RCA territories [PITH_FULL_IMAGE:figures/full_fig_p022_4.png]
Figure 5
Figure 5. Figure 5: Infarct territory clustering result of GMM-EM and CBMM-GICE ( [PITH_FULL_IMAGE:figures/full_fig_p023_5.png]
Figure 6
Figure 6. Figure 6: Clustering of LAD infarct patterns (N=54 patients with two visits, 2D visualization obtained from UMAP with KNN=6). Each dashed line links the two visits of the same patient (1 month (M1) and 12 months (M12) after the heart has been reperfused). In each subplot, the re…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 43 canonical work pages

  1. [1]

    Allaoui, M

    M. Allaoui, M. Kherfi, A. Cheriet, Considerably improving clustering al- gorithms using UMAP dimensionality reduction technique: A comparative study, in: Proc. International Conference on Image and Signal Processing, Springer, 317–325, 2020

    work page 2020

  2. [2]

    Banfield, A

    J. Banfield, A. Raftery, Model-based Gaussian and non-Gaussian cluster- ing, Biometrics 49 (1993) 803–21

    work page 1993

  3. [3]

    Baudry, A

    J. Baudry, A. Raftery, G. Celeux, K. Lo, R. Gottardo, Combining Mix- ture Components for Clustering*, Journal of Computational and Graphical Statistics 9 (2010) 332–53

    work page 2010

  4. [4]

    Bouguila, D

    N. Bouguila, D. Ziou, J. Vaillancourt, Unsupervised learning of a finite mixture model based on the Dirichlet distribution and its application, IEEE Transactions on Image Processing 13 (2004) 1533–1543

    work page 2004

  5. [5]

    Campello, D

    R. Campello, D. Moulavi, J. Sander, Density-based clustering based on hierarchical density estimates, in: Proc. Pacific-Asia Conference on Knowl- edge Discovery and Data Mining, Springer, 160–172, 2013. 2https://www.creatis.insa-lyon.fr/Challenge/myosaiq/ 28

    work page 2013

  6. [6]

    Cerqueira, N

    M. Cerqueira, N. Weissman, V. Dilsizian, A. Jacobs, S. Kaul, W. Laskey, D. Pennell, J. Rumberger, T. Ryan, M. Verani, Standardized myocardial segmentation and nomenclature for tomographic imaging of the heart, Jour- nal of the American Society of Echocardiography 15 (2002) 463–467

    work page 2002

  7. [7]

    Delignon, A

    Y. Delignon, A. Marzouki, W. Pieczynski, Estimation of generalized mix- tures and its application in image segmentation, IEEE Transactions on Image Processing 6 (1997) 1364–1375

    work page 1997

  8. [8]

    Derrode, W

    S. Derrode, W. Pieczynski, Unsupervised classification using hidden Markov chain with unknown noise copulas and margins, Signal Process- ing 128 (2016) 8–17

    work page 2016

Show all 43 references
  1. [9]

    Duchateau, M

    N. Duchateau, M. De Craene, M. Sitges, V. Caselles, Adaptation of multi- scale function extension to inexact matching: application to the mapping of individuals to a learnt manifold, Proc. Geometric Science of Information, LNCS 8085 (2013) 578–586

    2013

  2. [10]

    Duchateau, M

    N. Duchateau, M. Viallon, L. Petrusca, P. Clarysse, N. Mewton, L. Belle, P. Croisille, Pixel-wise statistical analysis of myocardial injury in STEMI patients with delayed enhancement MRI, Frontiers in Cardiovascular Medicine 10 (2023) 1136760

    2023

  3. [11]

    X. Fan, H. Joe, High-dimensional factor copula models with estimation of latent variables, Journal of Multivariate Analysis 201 (2024) 105263

    2024

  4. [12]

    Forbes, D

    F. Forbes, D. Wraith, A new family of multivariate heavy-tailed distribu- tions with variable marginal amounts of tailweight: application to robust clustering, Statistics and Computing 24 (2014) 971–984

    2014

  5. [13]

    Fujimaki, Y

    R. Fujimaki, Y. Sogawa, S. Morinaga, Online heterogeneous mixture mod- eling with marginal and copula selection, Proc. 17th ACM SIGKDD In- ternational Conference on Knowledge Discovery and Data Mining (2011) 645–653. 29

    2011

  6. [14]

    Genest, K

    C. Genest, K. Ghoudi, L. Rivest, A semiparametric estimation procedure of dependence parameters in multivariate families of distributions, Biometrika 82 (1995) 543–552

    1995

  7. [15]

    Hancer, D

    E. Hancer, D. Karaboga, A comprehensive survey of traditional, merge- split and evolutionary approaches proposed for determination of cluster number, Swarm and Evolutionary Computation 32 (2017) 49–67

    2017

  8. [16]

    Huang, S

    W. Huang, S. Dong, Probability distribution of wave periods in combined sea states with finite mixture models, Applied Ocean Research 92 (2019) 101938

    2019

  9. [17]

    Ibanez, S

    B. Ibanez, S. James, S. Agewall, M. Antunes, C. Bucciarelli-Ducci, H. Bueno, A. Caforio, F. Crea, J. Goudevenos, S. Halvorsen, G. Hindricks, A. Kastrati, M. Lenzen, E. Prescott, M. Roffi, M. Valgimigli, C. Varen- horst, P. Vranckx, P. Widimsk´ y, 2017 ESC Guidelines for the ma...

    2018

  10. [18]

    H. Joe, J. Xu, The estimation method of inference functions for margins for multivariate models, Technical Report, Department of Statistics, Uni- versity of British Columbia, Vancouver 166 (1996)

    1996

  11. [19]

    Kendall, A new measure of rank correlation, Biometrika 30 (1938) 81– 93

    M. Kendall, A new measure of rank correlation, Biometrika 30 (1938) 81– 93

    1938

  12. [20]

    Kosmidis, D

    I. Kosmidis, D. Karlis, Model-based clustering using copulas with applica- tions, Statistics and Computing 26 (2016) 1079–1099

    2016

  13. [21]

    Kuang, C

    X. Kuang, C. Yu, Generating cotton fiber length probability density func- tion from fiber length parameters by finite mixture model, The Journal of The Textile Institute 106 (2015) 655–662

    2015

  14. [22]

    Lecun, L

    Y. Lecun, L. Bottou, Y. Bengio, P. Haffner, Gradient-based learning ap- plied to document recognition, Proceedings of the IEEE 86 (1998) 2278– 2324. 30

    1998

  15. [23]

    S. Lee, G. McLachlan, Finite mixtures of multivariate skew t-distributions: some recent and new results, Statistics and Computing 24 (2014) 181–202

    2014

  16. [24]

    W. Li, J. Isberg, R. Waters, J. Engstr¨ om, O. Svensson, M. Leijon, Statis- tical analysis of wave climate data using mixed distributions and extreme wave prediction, Energies 9 (2016) 396

    2016

  17. [25]

    W. Li, X. Wang, G. Li, B. Geng, P. Varshney, NN-Copula-CD: A Copula- Guided Interpretable Neural Network for Change Detection in Heteroge- neous Remote Sensing Images, arXiv preprint arXiv:2303.17448

    arXiv

  18. [26]

    C. Liu, H. Li, K. Fu, F. Zhang, M. Datcu, W. Emery, Bayesian estimation of generalized gamma mixture model based on variational em algorithm, Pattern Recognition 87 (2019) 269–284

    2019

  19. [27]

    Z. Ma, A. Leijon, Bayesian estimation of beta mixture models with vari- ational inference, IEEE Transactions on Pattern Analysis and Machine Intelligence 33 (2011) 2160–2173

    2011

  20. [28]

    McConville, R

    R. McConville, R. Santos-Rodriguez, R. Piechocki, I. Craddock, N2d:(not too) deep clustering via clustering the local manifold of an autoencoded em- bedding, in: Proc. 25th International Conference on Pattern Recognition, IEEE, 5145–5152, 2021

    2021

  21. [29]

    McInnes, J

    L. McInnes, J. Healy, J. Melville, UMAP: Uniform manifold approximation and projection for dimension reduction, Journal of Open Source Software 3 (2018) 861

    2018

  22. [30]

    McLachlan, T

    G. McLachlan, T. Krishnan, The EM algorithm and extensions, vol. 382, John Wiley & Sons, 2007

    2007

  23. [31]

    McLachlan, S

    G. McLachlan, S. Lee, S. Rathnayake, Finite mixture models, Annual Re- view of Statistics and Its Application 6 (2019) 355–378

    2019

  24. [32]

    Nelsen, An introduction to copulas, Springer Science & Business Media, 2007

    R. Nelsen, An introduction to copulas, Springer Science & Business Media, 2007. 31

    2007

  25. [33]

    Pieczynski, Convergence of the iterative conditional estimation and ap- plication to mixture proportion identification, Proc

    W. Pieczynski, Convergence of the iterative conditional estimation and ap- plication to mixture proportion identification, Proc. IEEE/SP 14th Work- shop on Statistical Signal Processing (2007) 49–53

    2007

  26. [34]

    Y. Ren, J. Pu, Z. Yang, J. Xu, G. Li, X. Pu, P. Yu, L. He, Deep Clustering: A Comprehensive Survey, arXiv

  27. [35]

    Rousseeuw, Silhouettes: a graphical aid to the interpretation and vali- dation of cluster analysis, Journal of Computational and Applied Mathe- matics 20 (1987) 53–65

    P. Rousseeuw, Silhouettes: a graphical aid to the interpretation and vali- dation of cluster analysis, Journal of Computational and Applied Mathe- matics 20 (1987) 53–65

    1987

  28. [36]

    A. Roy, S. Parui, Pair-copula based mixture models and their application in clustering, Pattern Recognition 47 (2014) 1689–1697

    2014

  29. [37]

    Sahin, C

    ¨O. Sahin, C. Czado, Vine copula mixture models and clustering for non- Gaussian data, Econometrics and Statistics 22 (2022) 136–158

    2022

  30. [38]

    Sfikas, C

    G. Sfikas, C. Nikou, N. Galatsanos, Robust image segmentation with mix- tures of Student’s t-distributions, Proc. IEEE International Conference on Image Processing 1 (2007) I–273

    2007

  31. [39]

    Sklar, Fonctions de r´ epartition ` a n dimensions et leurs marges, Publi- cations de l’Institut Statistique de l’Universit´ e de Paris 8 (1959) 229–231

    M. Sklar, Fonctions de r´ epartition ` a n dimensions et leurs marges, Publi- cations de l’Institut Statistique de l’Universit´ e de Paris 8 (1959) 229–231

    1959

  32. [40]

    Tamborrino, A

    C. Tamborrino, A. Falini, F. Mazzia, Empirical Density Estimation based on Spline Quasi-Interpolation with applications to Copulas clustering mod- eling, arXiv preprint arXiv:2402.11552

    arXiv

  33. [41]

    Venna, S

    J. Venna, S. Kaski, Local multidimensional scaling with controlled trade- off between trustworthiness and continuity, Proc. 5th Workshop on Self- Organizing Maps (2005) 695–702

    2005

  34. [42]

    Zheng, S

    F. Zheng, S. Derrode, W. Pieczynski, Semi-supervised optimal recursive filtering and smoothing in non-Gaussian Markov switching models, Signal Processing 171 (2020) 107511. 32

    2020

  35. [43]

    loc”,“scale

    F. Zheng, M. Jalbert, F. Forbes, S. Bonnet, A. Wojtusciszyn, S. Lablanche, P. Benhamou, Characterization of daily glycemic variability in subjects with type 1 diabetes using a mixture of metrics, Diabetes Technology & Therapeutics 22 (2020) 301–313. 33 Appendix A: Marginals an...

    2020

Pith tools

Reviewed August 8, 2026 · model on record in the stance chip above.