Pith. sign in

REVIEW 2 major objections 7 minor 300 references

Parsimonious Mixtures of Skewed Bilinear Factor Analyzers

T0 review · 2 major / 7 minor · reviewed 2026-08-02 · deepseek-v4-flash

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

Pith's one-line read A family of 256 constrained mixture models for skewed matrix-variate data is introduced, with an AECM algorithm and evidence from simulated and real data.

desk verdict A useful but incremental extension of the authors' own PMMVBF A framework; the new A_g=A and ν_g=ν constraints give a 256-model family, but the Stage 3 constrained-Δ update is dimensionally wrong as printed and needs correction or code before the central claim is reproducible. read the letter →

arxiv 2607.14297 v1 pith:XPWTY32C submitted 2026-07-15 stat.ME stat.ML

Jacob Moore , Michael P.B. Gallaugher This is my paper
classification stat.MEstat.ML MSC 62H3062H25
keywords mixturemodelsmatrixvariatedistributionsskewtdistributionbilinearfactoranalyzersparsimoniousAECMalgorithmmodel-basedclusteringdimensionreduction
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper introduces a family of 256 mixture models for clustering skewed matrix-valued data, such as images or longitudinal records. Each model combines a matrix-variate skew-t component distribution with bilinear factor analysis and imposes constraints across clusters on factor loadings, scale matrices, skewness, and tail heaviness, trading flexibility for far fewer parameters. The authors derive an alternating expectation-conditional-maximization (AECM) algorithm to fit every member of the family and report simulation experiments where the true model is recovered with high frequency and parameter estimates improve as sample size grows. On subsets of the MNIST handwritten-digit data and the Olivetti faces data, the skewed parsimonious models improve adjusted Rand index compared with non-skewed counterparts. If correct, the work gives practitioners a drop-in toolbox for model-based clustering of matrix data with asymmetry and heavy tails.

What carries the argument

The central object is the PMSBFA family: each model is a mixture of G matrix-variate skew-t distributions, with each component written as X = M_g + W_g A_g + Λ_g U_g Δ_g' + noise. Parsimony is obtained by constraining, across clusters, the row factor loadings Λ_g, the column factor loadings Δ_g, the row noise scale Σ_g, the column noise scale Ψ_g (with options to make these isotropic or shared), the skewness matrix A_g, and the degrees of freedom ν_g. The estimation machinery is a three-stage AECM algorithm: stage 1 updates memberships, location, skewness, and ν; stage 2 updates Λ and Σ; stage 3 updates Δ and Ψ. The unique-solution property of the shared-skewness update is established by sho

What would settle it

Run the paper's simulation Study 1 for the CCC-CCC-C-C model with N=500 and 10×10 matrices, implementing the Stage 3 update exactly as printed. If the denominator in the shared-Δ update is indeed E^{(3)}_{1ig} (an n×r matrix) rather than E^{(3)}_{3ig} (an r×r matrix), the algorithm should fail or produce divergent estimates, contradicting the reported TMS=50; if the denominator is a typo corrected to E^{(3)}_{3ig}, the simulation should reproduce the paper's results.

Watch

Extended reading notes

Core claim

The central claim is that the constraining schemes previously used to make Gaussian mixture models parsimonious can be applied simultaneously to the row and column factor structures, the skewness matrix, and the degrees-of-freedom parameter of a matrix-variate skew-t mixture, yielding 256 distinct models—the Parsimonious Mixtures of Skewed Bilinear Factor Analyzers (PMSBFA) family. The paper proposes a three-stage AECM algorithm whose update equations depend on which of the eight row-model and eight column-model constraint patterns is active, and it provides a proof that the constrained skewness update has a unique solution at each iteration. The authors claim that in their simulations the f

Load-bearing premise

The load-bearing premise is that the printed AECM update equations are dimensionally correct for every one of the 256 constraint combinations; in particular, the Stage 3 update for the shared row-factor loading matrix when the column model is CCC, CCU, or CUC appears to divide by an n×r matrix where an r×r matrix is required, which would invalidate the M-step for the 128 models with a shared row-factor loading matrix if it is not a typographical error.

Editorial extensions

If this is right

  • If the estimation algorithm is correct, the 256-model family provides a practical model-selection grid: fitting all combinations and choosing by BIC or ICL identifies the right constraint structure without bespoke tuning.
  • Clustering of matrix data (images, repeated measures) no longer requires vectorization; the bilinear factor structure preserves row/column geometry while the skew-t component handles asymmetry and outliers.
  • The parsimonious constraints can reduce parameters per component from O(np + n^2 + p^2) to O(nq + pr + ...), making high-dimensional matrix clustering feasible.
  • The reported MNIST and Olivetti results suggest that including skewness and heavier tails improves cluster recovery over Gaussian parsimonious factor analyzers on real image data.
  • The semi-supervised extension means the same models can be used for classification when a fraction of labels is known.

Reading between the lines

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

  • Editorial inference: the simulation evidence is limited to four constraint patterns and low-dimensional matrices; the claim that the whole 256-model family is usable rests on the update equations being correct for all constraint combinations, which the paper does not verify empirically.
  • Editorial inference: the same constraint algebra could be applied to other skewed matrix-variate distributions (e.g., variance-gamma or generalized hyperbolic), potentially yielding analogous 256-model families.
  • Editorial inference: the reported ARI gains on MNIST (0.846 vs 0.713 for unsupervised clustering) are suggestive but come from a single two-digit subset; a broader benchmark would clarify when skewness pays off.
  • Editorial inference: the model-selection step fitting all 256 models for each (G, q, r) grid point is computationally heavy; the paper itself notes a search algorithm is a future direction, and such an algorithm could make the family practical at scale.
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

2 major / 7 minor

Summary. The paper introduces a 256-model family, denoted PMSBFA, for parsimonious mixtures of skewed matrix-variate bilinear factor analyzers in the skew-t case. Parsimony is obtained by applying the eight row/column constraints of Gallaugher and McNicholas to the two factor-analytic scale matrices and additionally allowing the skewness matrix and degrees of freedom to be shared across components. Estimation is carried out with a three-stage AECM algorithm, with special updates for constrained factor loadings and skewness. The paper reports simulation studies of model recovery and parameter estimation for four constrained models, and applies the family to MNIST and Olivetti faces data.

Significance. If the algorithm is correct, the PMSBFA family is a useful extension of existing parsimonious matrix-variate factor-analytic mixtures: it combines dimension reduction with parsimony across clusters and adds skewness and heavy tails, which is likely to be valuable in clustering matrix-valued data. The paper also contains a genuinely useful existence/uniqueness argument for the constrained skewness update in Section 4.1, and the real-data results suggest practical gains over the Gaussian PMMVBFA models. However, the central algorithmic claim is not supported as printed: the Stage 3 update for constrained row-factor loadings is dimensionally inconsistent for all models with Δ_g=Δ, which is half of the 256-model family and includes three of the four simulation designs. Because this is a correctable typographical or algebraic issue rather than a fundamental flaw, the appropriate action is major revision rather than rejection.

major comments (2)
  1. [Section 3.2, Stage 3 updates] The constrained-Δ updates for column models CCC, CCU, CUC, and CUU are dimensionally inconsistent as printed. For example, the CCC/CCU update divides by Σ_i Σ_g z_ig E^{(3)}_{1ig}, but E^{(3)}_{1ig} is n×r (as defined immediately above), while the inverse requires an r×r matrix. The natural correct denominator is E^{(3)}_{3ig}, which is r×r and is already used in the unconstrained-Δ update. The same substitution is needed in the CUC and CUU row-by-row updates. Unless this is a typographical error, the M-step is undefined for all 128 models with Δ_g=Δ, including simulation models C1–C3 in Table 5. The manuscript must either supply corrected formulas or clearly restrict the claims to models with unconstrained Δ.
  2. [Section 5 / Reproducibility] No code or supplementary material is provided, so the near-perfect model-recovery numbers in Table 5 for C1–C3 cannot be checked against the printed equations. Since those simulations rely on the invalid constrained-Δ updates, the current text does not allow a reader to reproduce the central empirical claim. I request corrected equations and, ideally, the code used for the simulations, or at least a precise numerical trace for one simulated dataset for each of C1–C4.
minor comments (7)
  1. [Abstract] The sentence 'In this manuscript propose a parsimonious family' is missing 'we'.
  2. [Section 2.4] 'first first discussed' should be 'first discussed'.
  3. [Section 2.4, Tables 1–2] The text refers to 'Table 2.4 and Table 2.4' instead of Table 1 and Table 2.
  4. [Section 3.2, Stage 3, CUC update] In the CUC constrained-Δ update, the term A'_g should presumably be \(\hat A'^{(t+1)}_g\) for consistency with the surrounding formulas.
  5. [Section 5.1] The text says 'the 24 total simulated populations' but the Cartesian product C × d has 4 × 3 = 12 populations. If the intention is to count both sample sizes separately, this should be stated explicitly.
  6. [Section 5 and 6] The notation is inconsistent: the model family is called PMSBFA in some places and PMSBF A in others. Please standardize.
  7. [Table 5 and Section 3.6] The table caption mentions 'the adaptive algorithm', but no adaptive model-selection algorithm is described in Section 3.6 or elsewhere. Please either describe it or remove the term.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the PMSBF A family and its constrained AECM updates are derived in-paper; self-citations supply background only.

full rationale

The paper's central contribution—the 256-model PMSBF A family and the AECM updates for constrained skewness and concentration—is not a disguised re-use of its inputs. The model family is generated by combining eight row/column constraints with A_g=A and ν_g=ν (Section 3.1), and the constrained skewness update (18) is derived from the Stage 1 complete-data log-likelihood (14); the positive-definiteness proof in Section 4.1 is completed in-text via Jensen's inequality rather than imported as an external 'uniqueness theorem'. The Stage 2 and Stage 3 factor-loading updates are stated from the complete-data log-likelihoods and parallel the unconstrained cases from Gallaugher and McNicholas (2020a), but the constrained cases are re-derived/adapted in the manuscript. The simulations generate data from PMSBF A populations and then check BIC recovery and parameter estimation; this is a standard estimation-consistency check with known ground truth, not a fitted value being relabeled as a prediction. Real-data analyses on MNIST and Olivetti use known class labels and provide external benchmarks. The self-citations to the authors' prior matrix-variate skew-t and bilinear factor analyzer papers supply the distributional background and the basic AECM structure, but they do not by themselves force the new constrained-model claims. One non-circular concern is that the Stage 3 constrained-Δ updates appear dimensionally inconsistent: the denominator is printed as E^{(3)}_{1ig} (n×r) rather than the r×r E^{(3)}_{3ig}; this is a correctness/reproducibility issue, not circularity. Because the central derivation is not equivalent to its inputs, no circular step is identified; score 2 reflects the paper's heavy but non-load-bearing reliance on self-cited machinery.

Assumptions & free parameters 1 free parameters · 3 assumptions · 0 invented entities

No entirely new latent entity is introduced; the model reuses the latent W, U, and E matrices from prior work. The central contribution is a constrained estimation scheme whose correctness inherits the stochastic representation and conditional distribution results of self-cited prior papers, and whose BIC-based model selection relies on identifiability and parameter-counting assumptions.

free parameters (1)
  • Initialization and tuning constants = 200 candidate starts; 20 kept; 4 small-EM iterations; A_g^(0)=0.1; ν^(0)=10; tolerance ϵ=|ℓ^(10)|×10^{-k} with k=log10(n
    Hand-chosen values that affect convergence and final solutions. The paper justifies them empirically but they are not derived from first principles.
assumptions (3)
  • domain assumption The matrix-variate skew-t density and its GIG conditional moments (Eqs. (3) and (6)) are correct for the variance-mean mixture representation X=M+WA+√W V with W∼IGamma(ν/2,ν/2).
    Eqs. (3)-(7) are taken from Gallaugher and McNicholas (2017, 2018a, 2019); the AECM E-step relies on these closed-form expectations. See Section 3.2.
  • domain assumption The factor decomposition V_ig=Λ_g U_ig Δ_g'+Λ_g E^B_ig+E^A_ig Δ_g'+E_ig with mutually independent error matrices and diagonal Σ_g, Ψ_g yields the marginal MVST components and the conditional linear expectations used in AECM stages 2 and 3.
    Eqs. (8)-(12); inherited from prior work. If this representation is not exact, the M-step updates are not valid.
  • domain assumption The parameter counts in Tables 1 and 2 are correct, including the rotation correction [nq−q(q−1)/2] for factor loadings, and identifiability can be fixed by restricting the first diagonal entry of Σ*_g or Ψ*.
    Section 3.4; BIC model selection (Eq. 21) uses these counts, so an error in counting would bias model choice.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Parsimonious Mixtures of Skewed Bilinear Factor Analyzers." pith.science (2026). https://pith.science/paper/XPWTY32C

@misc{pith2026260714297,
  author       = {Pith},
  title        = {Pith review of: Parsimonious Mixtures of Skewed Bilinear Factor Analyzers},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XPWTY32C}},
  note         = {Machine review of arXiv:2607.14297}
}
read the original abstract

Mixture models which cluster skewed random matrices can often suffer from over-parameterization in the absence of performing dimension reduction. Even with the use of bilinear factor analyzers, further parameter reduction can be achieved by constraining parameters over clusters. In this manuscript propose a parsimonious family of 256 models for mixtures of skewed matrix variate bilinear factor analyzers, specifically in the case of the skew t distribution. An AECM algorithm for parameter estimation is discussed in detail. Further, extensive simulations are performed, and the method is considered in the case of the MNIST dataset and the Olivetti faces dataset.

Figures

Figures reproduced from arXiv: 2607.14297 by the authors.

Figure 1
Figure 1. Depictions of the estimated mean fields for the PMSBFA model selected by the BIC. On the [PITH_FULL_IMAGE:figures/full_fig_p018_1.png] view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

300 extracted references

  1. [1]

    and Stiefel, Eduard , title =

    Hestenes, Magnus R. and Stiefel, Eduard , title =. Journal of Research of the National Bureau of Standards , year =

  2. [2]

    and Demmel, James and Donato, June and Dongarra, Jack and Eijkhout, Victor and Pozo, Roldan and Romine, Charles and van der Vorst, Henk , title =

    Barrett, Richard and Berry, Michael and Chan, Tony F. and Demmel, James and Donato, June and Dongarra, Jack and Eijkhout, Victor and Pozo, Roldan and Romine, Charles and van der Vorst, Henk , title =. 1994 , doi =

    1994

  3. [3]

    and Franczak, B

    Tortora, C. and Franczak, B. C. and Browne, R. P. and McNicholas, P. D. , title =. Journal of Classification , year =

  4. [4]

    Flexible clustering of high-dimensional data via mixtures of joint generalized hyperbolic distributions , journal =

    Tang, Y and Browne, R P and McNicholas, P D , year =. Flexible clustering of high-dimensional data via mixtures of joint generalized hyperbolic distributions , journal =

  5. [5]

    Murray, P. M. and Browne, R. B. and McNicholas, P. D. , Title =. Journal of Multivariate Analysis , Pages =. 2017 , Volume =

    2017

  6. [6]

    Murray, P. M. and Browne, R. B. and McNicholas, P. D. , Title =. Econometrics and Statistics , Pages =. 2017 , Volume =

    2017

  7. [7]

    and McLachlan, G

    Lin, T. and McLachlan, G. J. and Lee, S. X. , year =. Extending mixtures of factor models using the restricted multivariate skew-normal distribution , journal =

  8. [8]

    Advances in Data Analysis and Classification , volume=

    A mixture of generalized hyperbolic factor analyzers , author=. Advances in Data Analysis and Classification , volume=

Show all 300 references
  1. [9]

    Finite mixtures of matrix normal distributions for classifying three-way data , Volume =

    Viroli, Cinzia , Journal =. Finite mixtures of matrix normal distributions for classifying three-way data , Volume =

  2. [10]

    and McNicholas, Paul D

    McNicholas, Sharon M. and McNicholas, Paul D. and Browne, Ryan P. A Mixture of Variance-Gamma Factor Analyzers. Big and Complex Data Analysis: Methodologies and Applications. 2017

    2017

  3. [11]

    Statistics , volume=

    On the matrix-variate generalized hyperbolic distribution and its Bayesian applications , author=. Statistics , volume=. 2004 , publisher=

    2004

  4. [12]

    2013 , note =

    RnavGraphImageData: Some image data used in the RnavGraph package demos , author =. 2013 , note =

    2013

  5. [13]

    Proceedings of the IEEE , volume=

    Gradient-based learning applied to document recognition , author=. Proceedings of the IEEE , volume=. 1998 , publisher=

    1998

  6. [14]

    Pattern Recognition , volume=

    Finite mixtures of skewed matrix variate distributions , author=. Pattern Recognition , volume=. 2018 , publisher=

    2018

  7. [15]

    On model-based clustering of skewed matrix data , volume =

    Melnykov, Volodymyr and Zhu, Xuwen , year =. On model-based clustering of skewed matrix data , volume =

  8. [16]

    Proceedings of the Joint Statistical Meetings , venue =

    Mixtures of Matrix Variate Bilinear Factor Analyzers , author =. Proceedings of the Joint Statistical Meetings , venue =. 2018 , publisher =

    2018

  9. [17]

    IEEE Transactions on Neural Networks , volume=

    Matrix-variate factor analysis and its applications , author=. IEEE Transactions on Neural Networks , volume=. 2008 , publisher=

    2008

  10. [18]

    IEEE Transactions on Neural Networks and Learning Systems , volume=

    Bilinear probabilistic principal component analysis , author=. IEEE Transactions on Neural Networks and Learning Systems , volume=. 2012 , publisher=

    2012

  11. [19]

    Probabilistic interpretations and extensions for a family of

    Yu, Shipeng and Bi, Jinbo and Ye, Jieping , booktitle=. Probabilistic interpretations and extensions for a family of

  12. [20]

    and McNicholas, Paul D

    McNicholas, Sharon M. and McNicholas, Paul D. and Browne, Ryan P. , title =. Big and Complex Data Analysis, Contributions to Statistics , publisher =. 2017 , pages =

    2017

  13. [21]

    Stat , volume=

    A matrix variate skew-t distribution , author=. Stat , volume=. 2017 , publisher=

    2017

  14. [22]

    Michael P. B. Gallaugher and Paul D. McNicholas , Title =. 2019 , journal =

    2019

  15. [23]

    Gallaugher, Michael P. B. and McNicholas, Paul D. , title =. Advances in Data Analysis and Classification , volume =. 2020 , doi =

    2020

  16. [24]

    Michael P. B. Gallaugher and Paul D. McNicholas , title =. Advanced Studies in Behaviormetrics and Data Science , series =. 2020 , doi =

    2020

  17. [25]

    , title=

    Melnykov, V. , title=. 2016 , journal=

    2016

  18. [26]

    The Annals of Applied Statistics , volume=

    Covariance pattern mixture models for the analysis of multivariate heterogeneous longitudinal data , author=. The Annals of Applied Statistics , volume=. 2015 , publisher=

    2015

  19. [27]

    1999 , publisher=

    Matrix variate distributions , author=. 1999 , publisher=

    1999

  20. [28]

    Journal of statistical computation and simulation , volume=

    The MLE algorithm for the matrix normal distribution , author=. Journal of statistical computation and simulation , volume=. 1999 , publisher=

    1999

  21. [29]

    Statistics , volume=

    On matrix variate skew-normal distributions , author=. Statistics , volume=. 2008 , publisher=

    2008

  22. [30]

    Annals of Mathematical Statistics , volume=

    Estimating The Infinitesimal Generator of a Finite State Continuous-Time Markov Process , author=. Annals of Mathematical Statistics , volume=

  23. [31]

    Proceedings of the web mining workshop at the 1st SIAM conference on data mining , volume=

    Clickstream clustering using weighted longest common subsequences , author=. Proceedings of the web mining workshop at the 1st SIAM conference on data mining , volume=. 2001 , organization=

    2001

  24. [32]

    2000 , publisher=

    Concept-based clustering of clickstream data , author=. 2000 , publisher=

    2000

  25. [33]

    Data Mining and Knowledge Discovery , volume=

    Model-based clustering and visualization of navigation patterns on a web site , author=. Data Mining and Knowledge Discovery , volume=. 2003 , publisher=

    2003

  26. [34]

    Marketing Science , volume=

    Modeling online browsing and path analysis using clickstream data , author=. Marketing Science , volume=. 2004 , publisher=

    2004

  27. [35]

    Visual Analytics Science and Technology (VAST), 2012 IEEE Conference on , pages=

    Visual cluster exploration of web clickstream data , author=. Visual Analytics Science and Technology (VAST), 2012 IEEE Conference on , pages=. 2012 , organization=

    2012

  28. [36]

    Proceedings of the 29th international conference on Very large data bases-Volume 29 , pages=

    A framework for clustering evolving data streams , author=. Proceedings of the 29th international conference on Very large data bases-Volume 29 , pages=. 2003 , organization=

    2003

  29. [37]

    , year =

    Akaike, H. , year =. A new look at the statistical model identification , journal =

  30. [38]

    Decision boundaries for mixtures of regressions , Year =

    Ingrassia, Salvatore and Punzo, Antonio , Journal =. Decision boundaries for mixtures of regressions , Year =

  31. [39]

    Advances in Statistical Models for Data Analysis , series =

    Punzo, Antonio and Ingrassia, Salvatore , title =. Advances in Statistical Models for Data Analysis , series =

  32. [40]

    McNicholas, P. D. , title =. Statistical Models for Data Analysis , series =

  33. [41]

    Variational B ayesian inference for infinite generalized inverted D irichlet mixtures with feature selection and its application to clustering

    Bdiri, Taoufik and Bouguila, Nizar and Ziou, Djemel. Variational B ayesian inference for infinite generalized inverted D irichlet mixtures with feature selection and its application to clustering. Applied Intelligence. 2016

    2016

  34. [42]

    Journal of Computational and Graphical Statistics , volume =

    Sonia Jain and Radford M Neal , title =. Journal of Computational and Graphical Statistics , volume =

  35. [43]

    Brusco , title =

    Emilie Shireman and Douglas Steinley and Michael J. Brusco , title =. Behavior Research Methods , Year =

  36. [44]

    Rao, C. R. , title =. 1952 , address =

    1952

  37. [45]

    , title =

    Pearson, K. , title =. Philosophical Transactions of the Royal Society, Part A , Volume =

  38. [46]

    and Gormley, I

    McParland, D. and Gormley, I. C. and McCormick, T. H. and Clark, S. J. and Kabudula, C. W. and Collinson, M. A. , Journal =. CLUSTERING SOUTH AFRICAN HOUSEHOLDS BASED ON THEIR ASSET STATUS USING LATENT VARIABLE MODELS , Volume =

  39. [47]

    Coretto and C

    P. Coretto and C. Hennig , note =. Robust improper maximum likelihood: tuning, computation, and a comparison with other methods for robust

  40. [48]

    , Journal =

    Viroli, C. , Journal =. Dimensionally Reduced Model-Based Clustering Through Mixtures of Factor Mixture Analyzers , Volume =

  41. [49]

    , Journal =

    Baudry, J.-P. , Journal =. Estimation and model selection for model-based clustering with the conditional classification likelihood , Volume =

  42. [50]

    What are the true clusters? , Volume =

    Hennig, C , Journal =. What are the true clusters? , Volume =

  43. [51]

    McQuitty, L. L. , Journal =. Agreement analysis: A method of classifying subjects according to their patterns of responses , Volume =

  44. [52]

    , Title =

    Stephensen, W. , Title =. 1953 , address =

    1953

  45. [53]

    Catell, R. B. , Journal =

  46. [54]

    Tryon, R. C. , year =. Identification of social areas by cluster analysis , booktitle =

  47. [55]

    and Biernacki, C

    Marbac, M. and Biernacki, C. and Vandewalle, V. , title =. 2015 , note =

    2015

  48. [56]

    Greselin and Salvatore Ingrassia , Journal =

    F. Greselin and Salvatore Ingrassia , Journal =. Constrained monotone

  49. [57]

    and Papla, D

    Jajuga, K. and Papla, D. , title =. 2006 , booktitle =

    2006

  50. [58]

    Di Lascio, F. M. L. and Giannerini, S. , year =. A copula-based algorithm for discovering patterns of dependent observations , journal =

  51. [59]

    Montanari and C

    A. Montanari and C. Viroli , year =. Maximum likelihood estimation of Mixture of Factor Analyzers , journal =

  52. [60]

    Montanari and C

    A. Montanari and C. Viroli , year =. Heteroscedastic Factor Mixture Analysis , journal =

  53. [61]

    Vrac, M. and L. Billard and E. Diday and A. Chedin , year =. Copula analysis of mixture models , journal =

  54. [62]

    Kosmidis, I. and D. Karlis , year =. Model-based clustering using copulas with applications , note =

  55. [63]

    Montanari and C

    A. Montanari and C. Viroli , year =. A skew-normal factor model for the analysis of student satisfaction towards university courses , journal =

  56. [64]

    Galimberti and A

    G. Galimberti and A. Montanari and C. Viroli , year =. Penalized Factor Mixture Analysis for Variable Selection in Clustered Data , journal =

  57. [65]

    McNicholas, P. D. , title =. 2016 , publisher =

    2016

  58. [66]

    McNicholas, P. D. , title =. 2016 , journal =

    2016

  59. [67]

    1981 , publisher=

    Finite Mixture Distributions , author=. 1981 , publisher=

    1981

  60. [68]

    Aitken, A. C. , Title =. Proceedings of the Royal Society of Edinburgh , Volume =

  61. [69]

    and Brusco, M

    Steinley, D. and Brusco, M. J. , year =. Selection of variables in cluster analysis: An empirical comparison of eight procedures , journal =

  62. [70]

    and McNicholas, P

    Vrbik, I. and McNicholas, P. D. , year =. Fractionally-supervised classification , journal =

  63. [71]

    Calibration of Multivariate Generalized Hyperbolic Distributions Using the

    Hu, Wenbo , year=. Calibration of Multivariate Generalized Hyperbolic Distributions Using the

  64. [72]

    , TITLE =

    Celeux, Gilles and Martin-Magniette, Marie-Laure and Maugis-Rabusseau, Cathy and Raftery, Adrian E. , TITLE =. Journal de la Soc\' et\'e Fran caise de Statistique , VOLUME =. 2014 , NUMBER =

    2014

  65. [73]

    and Sedki, M

    Marbac, M. and Sedki, M. , title =. 2015 , note =

    2015

  66. [74]

    Biernacki and G

    C. Biernacki and G. Celeux and G. Govaert , year =. Choosing Starting Values for the. Computational Statistics and Data Analysis , volume =

  67. [75]

    International Statistical Review , volume =

    The t Copula and Related Copulas , author =. International Statistical Review , volume =. 2005 , pages =

    2005

  68. [76]

    and Sha, Naijun and Vannucci, Marina , TITLE =

    Tadesse, Mahlet G. and Sha, Naijun and Vannucci, Marina , TITLE =. Journal of the American Statistical Association , VOLUME =. 2005 , NUMBER =

    2005

  69. [77]

    IEEE Transactions on Pattern Analysis and Machine Intelligence , volume=

    Simultaneous feature selection and clustering using mixture models , author=. IEEE Transactions on Pattern Analysis and Machine Intelligence , volume=

  70. [78]

    2016 , journal =

    Hypothesis testing for mixture model selection , author =. 2016 , journal =

    2016

  71. [79]

    2015 , volume=

    Robust Cluster Analysis and Variable Selection , author=. 2015 , volume=

    2015

  72. [80]

    2015 , note =

    ContaminatedMixt: Model-Based Clustering and Classification with the Multivariate Contaminated Normal Distribution , author =. 2015 , note =

    2015

  73. [81]

    Some methods for classification and analysis of multivariate observations

    MacQueen, J. Some methods for classification and analysis of multivariate observations. Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability, Volume 1: Statistics. 1967

    1967

  74. [82]

    , year =

    Steinley, D. , year =. K-means clustering: A half-century synthesis , journal =

  75. [83]

    Fabrigar and Duane T

    Leandre R. Fabrigar and Duane T. Wegener and Robert C. MacCallum and Erin J. Strahan , journal =. 1999 , volume =

    1999

  76. [84]

    Horn, J. L. , year =. A rationale and technique for estimating the number of factors in factor analysis , journal =

  77. [85]

    Humphreys, L. G. and Ilgen, D. R. , year =. Note on a criterion for the number of common factors , journal =

  78. [86]

    Humphreys, L. G. and Montanelli, Jr., R. G. , year =. An investigation of the parallel analysis criterion for determining the number of common factors , journal =

  79. [87]

    G and Humphreys, L

    Montanelli, Jr., R. G and Humphreys, L. G. , year =. Latent roots of random data correlation matrices with squared multiple correlations on the diagonal: A. Psychometrika , volume =

  80. [88]

    Zhou, H. and K. L. Lange , year =. On the bumpy road to the dominant mode , journal =

  81. [89]

    Advances in Neural Information Processing Systems , volume =

    Hagai Attias , title =. Advances in Neural Information Processing Systems , volume =. 2000 , pages =

    2000

  82. [90]

    Tipping, M. E. , year =. Probabilistic visualization of high-dimensional binary data , journal =

  83. [91]

    Jaakkola, T. S. and M. I. Jordan , year =. Bayesian parameter estimation via variational methods , journal =

  84. [92]

    and Viroli, C

    Cagnone, S. and Viroli, C. , year =. A factor mixture analysis model for multivariate binary data , journal =

  85. [93]

    Multilevel mixture item response theory models: an application in education testing , booktitle =

    Vermunt, J K , year =. Multilevel mixture item response theory models: an application in education testing , booktitle =

  86. [94]

    Item response mixture modeling: Application to tobacco dependence criteria , journal =

    Muthen, B and Asparouhov, T , year =. Item response mixture modeling: Application to tobacco dependence criteria , journal =

  87. [95]

    Mixture of latent trait analyzers for model-based clustering of categorical data , journal =

    Gollini, I and Murphy, T B , year =. Mixture of latent trait analyzers for model-based clustering of categorical data , journal =

  88. [96]

    , year =

    Goodman, L. , year =. Exploratory latent structure analysis using both identifiable and unidentifiable models , journal =

  89. [97]

    Celeux, G. and G. Govaert , year =. Clustering criteria for discrete data and latent class models , journal =

  90. [98]

    Biernacki, C. and G. Celeux and G. Govaert , year =. Exact and. Journal of Statistical Planning and Inference , volume =

  91. [99]

    Vermunt, J. K. , year =. Multilevel latent class models , journal =

  92. [100]

    Marbac, M. and C. Biernacki and V. Vandewalle , year =. Finite mixture model of conditional dependencies modes to cluster categorical data , note =

  93. [101]

    M. Lichman. UCI Machine Learning Repository. 2013

    2013

  94. [102]

    Churchill by Himself: The Definitive Collection of Quotations , year =

  95. [103]

    Campbell, J. G. and Fraley, C. and Murtagh, F. and Raftery, A. E. Linear flaw detection in woven textiles using model-based clustering. Pattern Recognition Letters. 1997

    1997

  96. [104]

    and Murphy, T

    O'Hagan, A. and Murphy, T. B. and Gormley, I. C. and McNicholas, P. D. and Karlis, D. , year =. Clustering with the multivariate normal inverse. Computational Statistics and Data Analysis , volume =

  97. [105]

    and Browne, R

    Tang, Y. and Browne, R. P. and McNicholas, P. D. , year =. Model-based clustering of high-dimensional binary data , journal =

  98. [106]

    Bioinformatics , year =

    Andrea Rau and Cathy Maugis-Rabusseau and Marie-Laure Martin-Magniette and Gilles Celeux , title =. Bioinformatics , year =

  99. [107]

    2008 , author =

    Model-based clustering for longitudinal data , journal =. 2008 , author =

    2008

  100. [108]

    , year =

    Melnykov, V. , year =. Model-Based Biclustering of Clickstream Data , journal =

  101. [109]

    Steele, R. J. and Raftery, A. E. , year =. Performance of. Frontiers of Statistical Decision Making and Bayesian Analysis , volume =

  102. [110]

    Psychological Methods , year =

    Evaluating Mixture Modeling for Clustering: Recommendations and Cautions , author =. Psychological Methods , year =

  103. [111]

    Psychological Methods , year =

  104. [112]

    Cluster Analysis , publisher =

    B S Everitt and S Landau and M Leese and D Stahl , year =. Cluster Analysis , publisher =

  105. [113]

    and Subedi, S

    Humbert, S. and Subedi, S. and Cohn, J. and Zeng, B. and Bi, Y.-M. and Chen, X. and Zhu, T. and McNicholas, P. D. and Rothstein, S. J. , year =. Genome-wide expression profiling of maize in response to individual and combined water and nitrogen stresses , journal =

  106. [114]

    and McLachlan, G

    Lee, S. and McLachlan, G. J. , year =. Finite mixtures of multivariate skew t-distributions: some recent and new results , journal =

  107. [115]

    Sahu and D

    K. Sahu and D. K. Dey and M. D. Branco , Journal =. A new class of multivariate skew distributions with applications to

  108. [116]

    Pyne and X

    S. Pyne and X. Hu and K. Wang and E. Rossin and T.-I. Lin and L. M. Maier and C. Baecher-Allan and G. J. McLachlan and P. Tamayo and D. A. Hafler and P. L. De Jager and J.P Mesirow , Journal =. Automated high-dimensional flow cytometric data analysis , Volume =

  109. [117]

    M. D. Branco and D. K. Dey , Journal =. A general class of multivariate skew-elliptical distributions , Volume =

  110. [118]

    The Skew-Normal and Related Families , Year =

    Adelchi Azzalini and Antonella Capitanio , Publisher =. The Skew-Normal and Related Families , Year =

  111. [119]

    Azzalini and A

    A. Azzalini and A. Dalla Valle , Journal =. The multivariate skew-normal distribution , Volume =

  112. [120]

    Azzalini and A

    A. Azzalini and A. Capitanio , Journal =. Statistical applications of the multivariate skew normal distribution , Volume =

  113. [121]

    Azzalini and A

    A. Azzalini and A. Capitanio , Journal =. Distributions generated by perturbation of symmetry with emphasis on a multivariate skew t distribution , Volume =

  114. [122]

    1999 , author=

    Molecular classification of cancer: Class discovery and class prediction by gene expression monitoring , journal=. 1999 , author=

    1999

  115. [123]

    Hunter, D. R. and Lange, K. , year =. Quantile regression via an. Journal of Computational and Graphical Statistics , volume =

  116. [124]

    Boik, Robert. J. , year =. Spectral Models for Covariance Matrices , journal =

  117. [125]

    Flury, B. W. and Gautschi, W. , year =. An algorithm for simultaneous orthogonal transformation of several positive definite symmetric matrices to nearly diagonal form , journal =

  118. [126]

    Lefkomtch, L. P. , year =. Consensus Principal Components , journal =

  119. [127]

    and McNicholas, P

    Subedi, S. and McNicholas, P. D. , Journal =. Variational

  120. [128]

    , year =

    Schott, J.R. , year =. Estimating correlation matrices that have common eigenvectors , journal =

  121. [129]

    Spiegelhalter, D. J. and Best, N. G. and Carlin, B. P. and Van der Linde, A. , Title =. Journal of the Royal Statistical Society: Series B , Volume =

  122. [130]

    and Ghahramani, Z

    Ueda, N. and Ghahramani, Z. , title =. Neural Networks , volume =

  123. [131]

    and Hurn, M

    Celeux, G. and Hurn, M. and Robert, C. P. , TITLE=. Journal of the American Statistical Association , VOLUME=

  124. [132]

    Variational Algorithms for Approximate

    Beal, Matthew James , year=. Variational Algorithms for Approximate

  125. [133]

    McGrory, C. A. and Titterington, D. M. , title =. Australian and New Zealand Journal of Statistics , volume =

  126. [134]

    McGrory, C. A. and Titterington, D. M. and Pettitt, A. N. , title =. Computational Statistics and Data Analysis , volume =

  127. [135]

    and Holmes, C

    Jasra, A. and Holmes, C. C. and Stephens, D. A. , year =. Markov Chain. Journal of the Royal Statistical Society: Series B , volume =

  128. [136]

    and Robert, C.P

    Diebolt, J. and Robert, C.P. , Journal =. Estimation of finite mixture distributions through. 1994 , Volume =

    1994

  129. [137]

    and Mengersen, K

    Casella, G. and Mengersen, K. L. and Robert, C. P. and Titterington, D. M. , title =. Journal of the Royal Statistical Society: Series B , volume =

  130. [138]

    , Title =

    Stephens, M. , Title =

  131. [139]

    Jordan, M. I. and Ghahramani, Z. and Jaakkola, T. S. and Saul, L. K. , title =. 1999 , journal =

    1999

  132. [140]

    and Wang, Y

    Teschendorff, A. and Wang, Y. and Barbosa-Morais, N. Brenton,J. and Caldas, C. A variational. Bioinformatics , Volume =

  133. [141]

    Journal of Computational and Graphical Statistics , year=

    Clustering Based on a Multi-layer Mixture Model , author=. Journal of Computational and Graphical Statistics , year=

  134. [142]

    2016 , journal =

    On nomenclature for, and the relative merits of, two formulations of skew distributions , author =. 2016 , journal =

    2016

  135. [143]

    McNicholas , year =

    Sanjeena Subedi and Paul D. McNicholas , year =. A Variational Approximations-

  136. [144]

    McNicholas , year =

    Antonio Punzo and Paul D. McNicholas , year =. Robust High-Dimensional Modeling with the Contaminated

  137. [145]

    McNicholas , year =

    Antonio Punzo and Paul D. McNicholas , year =. Robust Clustering in Regression Analysis via the Contaminated

  138. [146]

    2014 , note =

    Robust Asymmetric Clustering , author =. 2014 , note =

    2014

  139. [147]

    2015 , note =

    Clustering bivariate mixed-type data via the cluster-weighted model , author =. 2015 , note =

    2015

  140. [148]

    McNicholas , year =

    Antonio Punzo and Paul D. McNicholas , year =. Parsimonious mixtures of contaminated

  141. [149]

    Heinrich Fritz and Luis A. Garc\'. Journal of Statistical Software , year =

  142. [150]

    The Annals of Statistics , volume =

    A General Trimming Approach to Robust Cluster Analysis , author =. The Annals of Statistics , volume =

  143. [151]

    The Annals of Statistics , volume =

    A constrained formulation of maximum likelihood estimation for normal mixture distributions , author =. The Annals of Statistics , volume =

  144. [152]

    , year =

    Kharin, Y. , year =. Robustness in Statistical Pattern Recognition , publisher =

  145. [153]

    Huber, P. J. , year =. Robust estimation of a location parameter , journal =

  146. [154]

    Huber, P. J. , year =. Robust Statistics , publisher =

  147. [155]

    The Annals of Statistics , volume=

    Breakdown points for maximum likelihood estimators of location-scale mixtures , author=. The Annals of Statistics , volume=

  148. [156]

    and Vesin, J

    Schroeter, P. and Vesin, J. and Langenberger, T. and Meuli, R. , title=. IEEE Transactions on Medical Imaging , year=

  149. [157]

    Biometrics , volume =

    Markatou, Marianthi , title =. Biometrics , volume =

  150. [158]

    Statistics and Probability Letters , volume=

    Robust estimation of a normal mixture , author=. Statistics and Probability Letters , volume=

  151. [159]

    Mathematical Geology , volume=

    Mixture models and atypical values , author=. Mathematical Geology , volume=

  152. [160]

    Mixture models, outliers, and the

    Aitkin, Murray and Wilson, Granville Tunnicliffe , journal=. Mixture models, outliers, and the

  153. [161]

    Czerniak and H

    J. Czerniak and H. Zarzycki , title =. 2003 , booktitle =

    2003

  154. [162]

    Journal of the Royal Statistical Society: Series A , volume =

    Statistical Modelling of Data on Teaching Styles , author =. Journal of the Royal Statistical Society: Series A , volume =

  155. [163]

    Marriott, F. H. C. , year =. The Interpretation of Multiple Observations , publisher =

  156. [164]

    Hadley Wickham , title =

  157. [165]

    and Asuncion, A

    Frank, A. and Asuncion, A. , year =

  158. [166]

    Covariance pattern mixture models for multivariate longitudinal data , journal =

    Laura Anderlucci and Cinzia Viroli , year =. Covariance pattern mixture models for multivariate longitudinal data , journal =

  159. [167]

    and Celeux, G

    Bouveyron, C. and Celeux, G. and Girard, S. , year =. Intrinsic dimension estimation by maximum likelihood in isotropic probabilistic. Pattern Recognition Letters , volume =

  160. [168]

    2014 , month =

    Mixtures of Multiple Scaled Generalized Hyperbolic Distributions , author =. 2014 , month =

    2014

  161. [169]

    and McNicholas, P

    Morris, K. and McNicholas, P. D. , title =. Computational Statistics and Data Analysis , volume =

  162. [170]

    and McNicholas, P

    Morris, K. and McNicholas, P. D. and Scrucca, L. , Title =. Advances in Data Analysis and Classification , Pages =

  163. [171]

    and McNicholas, P

    Morris, K. and McNicholas, P. D. , Title =. Statistics and Probability Letters , Pages =

  164. [172]

    , year =

    Scrucca, L. , year =. Graphical tools for model-based mixture discriminant analysis , journal =

  165. [173]

    Model-based cluster and discriminant analysis with the

    Biernacki, Christophe and Celeux, Gilles and Govaert, G. Model-based cluster and discriminant analysis with the. Computational Statistics and Data Analysis , volume =. 2006 , pages =

    2006

  166. [174]

    Li, K. C. , title =. Journal of the American Statistical Association , pages =

  167. [175]

    Li, K. C. , title =. 2000 , note =

    2000

  168. [176]

    , year =

    Maugis, C. , year =. The selvarclust software , note =

  169. [177]

    2014 , journal =

    Model-based clustering of high-dimensional data: A review , author =. 2014 , journal =

    2014

  170. [178]

    Thorndike, R. L. , year =. Who belongs in the family? , journal =

  171. [179]

    Sanford Weisberg , Note =

  172. [180]

    Applied Linear Regression , edition =

    Sanford Weisberg , year =. Applied Linear Regression , edition =

  173. [181]

    and Celeux, G

    Bensmail, H. and Celeux, G. , TITLE =. Journal of the American Statistical Association , YEAR =

  174. [182]

    Altman, E. I. , year =. Financial ratios, discriminant analysis and the prediction of corporate bankruptcy , journal =

  175. [183]

    and Forbes, F

    Wraith, D. and Forbes, F. , Note =. Clustering using skewed multivariate heavy tailed distributions with flexible tail behaviour , Year =

  176. [184]

    Franczak and Cristina Tortora and Ryan P

    Brian C. Franczak and Cristina Tortora and Ryan P. Browne and Paul D. McNicholas , Journal =. Unsupervised learning via mixtures of skewed distributions with hypercube contours , Volume =

  177. [185]

    Normal scale mixtures and dual probability densities , Volume =

    Gneiting, Tilmann , Journal =. Normal scale mixtures and dual probability densities , Volume =

  178. [186]

    A new family of multivariate heavy-tailed distributions with variable marginal amounts of tailweights: Application to robust clustering , Volume =

    Forbes, Florence and Wraith, Darren , Journal =. A new family of multivariate heavy-tailed distributions with variable marginal amounts of tailweights: Application to robust clustering , Volume =

  179. [187]

    Barndorff-Nielsen and J

    O. Barndorff-Nielsen and J. Kent and M. S. Normal Variance-Mean Mixtures and z Distributions , Volume =. International Statistical Review , Number =

  180. [188]

    and van de Rijn, Matt and Botstein, David and Brown, Patrick O

    Chen, Xin and Cheung, Siu Tim and So, Samuel and Fan, Sheung Tat and Barry, Christopher and Higgins, John and Lai, Kin-Man and Ji, Jiafu and Dudoit, Sandrine and Ng, Irene O.L. and van de Rijn, Matt and Botstein, David and Brown, Patrick O. , Journal =. Gene Expression Pattern...

  181. [189]

    Multivariate t Distributions and Their Applications , Year =

    Kotz, Samuel and Nadarajah, Saralees , Date-Added =. Multivariate t Distributions and Their Applications , Year =

  182. [190]

    and Kanehisa, M

    Nakai, K. and Kanehisa, M. , Journal =. Expert System for Predicting Protein Localization Sites in Gram-Negative Bacteria , Volume =

  183. [191]

    and Kanehisa, M

    Nakai, K. and Kanehisa, M. , Journal =. A knowledge base for predicting protein localization sites in eukaryotic cells , Volume =

  184. [192]

    M. M. Ali and J. Woo and S. Nadarajah , Journal =. Some skew symmetric inverse reflected distributions , Volume =

  185. [193]

    2015 , note =

    MixGHD: Model Based Clustering, Classification and Discriminant Analysis Using the Mixture of Generalized Hyperbolic Distributions , author =. 2015 , note =

    2015

  186. [194]

    Tortora and P

    C. Tortora and P. D. McNicholas and R. P. Browne , Journal =. A Mixture of Generalized Hyperbolic Factor Analyzers , Year =

  187. [195]

    A Mixture of Coalesced Generalized Hyperbolic Distributions , Author =

  188. [196]

    1951 , address =

    Science: Sense and Nonsense , author =. 1951 , address =

    1951

  189. [197]

    2012 , note =

    Flury: Data Sets from Flury, 1997 , author =. 2012 , note =

    1997

  190. [198]

    Modern Applied Statistics with S , author =

  191. [199]

    G. M. Reaven and R. G. Miller , journal =. An attempt to define the nature of chemical diabetes using a multidimensional analysis , volume =

  192. [200]

    and Peterson, L

    Heinz, G. and Peterson, L. J. and Johnson, R. W. and Kerk, C. J. , year =. Exploring Relationships in Body Dimensions , journal =

  193. [201]

    Browne, R. P. and McNicholas, P. D. , year =. Orthogonal. Statistics and Computing , volume =

  194. [202]

    Fowlkes, E. B. and Mallows, C. L. , year =. A method for comparing two hierarchical clusterings , journal =

  195. [203]

    and Brennan, T

    Downton, M. and Brennan, T. , year =. Comparing classifications: An evaluation of several coefficients of partition agreement. , note =

  196. [204]

    Hubert, L. J. , year =. Nominal Scale Response Agreement as a Generalized Correlation , journal =

  197. [205]

    and Boorman, S

    Arabie, P. and Boorman, S. A , year =. Multidimensional Scaling of Measures of Distance Between Partitions , journal =

  198. [206]

    Mirkin, B. G. and Chernyi, L. B. , year =. Measurement of the Distance Between Distinct Partitions of a Finite Set of Objects , journal =

  199. [207]

    and Agresti, A

    Morey, L. and Agresti, A. , year =. The measurement of classification agreement: An adjustment to the. Educational and Psychological Measurement , volume =

  200. [208]

    and Van Den Berg, Robert A

    Van Deun, Katrijn and Wilderjans, Torn F. and Van Den Berg, Robert A. and Antoniadis, Anestis and Van Mechelen, I. , Journal =. A flexible framework for sparse simultaneous component based data integration , Year =

  201. [209]

    and Hunter, D

    Lange, K. and Hunter, D. R. and Yang, I. , Journal =. Optimization transfer using surrogate objective functions , Year =

  202. [210]

    , editor=

    De Leeuw, J. , editor=. Information Systems and Data Analysis , Pages =. Block Relaxation Algorithms in Statistics , Year =

  203. [211]

    Heiser, W. J. , editor =. Convergent Computation by Iterative Majorization: Theory and Applications in Multidimensional Data Analysis , title =. 1995 , address =

    1995

  204. [212]

    Journal of the American Statistical Association , VOLUME =

    Fan, Jianqing and Li, Runze , TITLE =. Journal of the American Statistical Association , VOLUME =. 2001 , PAGES =

    2001

  205. [213]

    and McNicholas, P

    Bhattacharya, S. and McNicholas, P. D. , Title =

  206. [214]

    and McNicholas, P

    Bhattacharya, S. and McNicholas, P. D. , Journal =. A

  207. [215]

    Proceedings of the National Academy of Sciences of the United States of America , volume =

    Broad Patterns of Gene Expression Revealed by Clustering Analysis of Tumor and Normal Colon Tissues Probed by Oligonucleotide Arrays , author =. Proceedings of the National Academy of Sciences of the United States of America , volume =. 1999 , publisher =

    1999

  208. [216]

    McLachlan, G. J. and Bean, R. W. and Peel, D. A mixture model-based approach to the clustering of microarray expression data , Journal =

  209. [217]

    Proceedings of the National Academy of Sciences of the United States of America , volume =

    Coupled Two-Way Clustering Analysis of Gene Microarray Data , author =. Proceedings of the National Academy of Sciences of the United States of America , volume =

  210. [218]

    David Meyer and Evgenia Dimitriadou and Kurt Hornik and Andreas Weingessel and Friedrich Leisch , year =

  211. [219]

    Journal of the American Statistical Association , volume=

    A framework for feature selection in clustering , author=. Journal of the American Statistical Association , volume=

  212. [220]

    Witten and Robert Tibshirani , year =

    Daniela M. Witten and Robert Tibshirani , year =

  213. [221]

    2012 , volume =

    Laurent Berg\'e and Charles Bouveyron and St\'ephane Girard , journal =. 2012 , volume =

    2012

  214. [222]

    Laurent Berg\'e and Charles Bouveyron and St\'ephane Girard , year =

  215. [223]

    Andrews and Jaymeson R

    Jeffrey L. Andrews and Jaymeson R. Wickins and Nicholas M. Boers and Paul D. McNicholas , year =

  216. [224]

    Andrews and Paul D

    Jeffrey L. Andrews and Paul D. McNicholas , year =

  217. [225]

    Mixtures of Multivariate Power Exponential Distributions , journal =

    U J Dang and R P Browne and P D McNicholas , year =. Mixtures of Multivariate Power Exponential Distributions , journal =

  218. [226]

    Variable selection for clustering and classification , journal =

    J L Andrews and P D McNicholas , year =. Variable selection for clustering and classification , journal =

  219. [227]

    North American Actuarial Journal , volume=

    Tail conditional expectations for elliptical distributions , author=. North American Actuarial Journal , volume=

  220. [228]

    1980 , publisher=

    Multivariate Analysis , author=. 1980 , publisher=

    1980

  221. [229]

    2008 , publisher=

    Optimization Algorithms on Matrix Manifolds , author=. 2008 , publisher=

    2008

  222. [230]

    Multivariate exponential power distributions as mixtures of normal distributions with

    G. Multivariate exponential power distributions as mixtures of normal distributions with. Communications in Statistics -- Theory and Methods , volume=

  223. [231]

    Multivariate Generalized

    Zhang, Teng and Wiesel, Ami and Grec, Maria Sabrina , journal=. Multivariate Generalized. 2013 , month=

    2013

  224. [232]

    IEEE Transactions on Signal Processing , title=

    Pascal, Fr. IEEE Transactions on Signal Processing , title=. 2013 , volume=

    2013

  225. [233]

    Biometrics , volume =

    Zhang, Jian and Liang, Faming , title =. Biometrics , volume =

  226. [234]

    Multiscale colour texture retrieval using the geodesic distance between multivariate generalized

    Verdoolaege, Geert and De Backer, Steve and Scheunders, Paul , booktitle=. Multiscale colour texture retrieval using the geodesic distance between multivariate generalized

  227. [235]

    Signal Processing: Image Communication , volume=

    Multivariate statistical modeling for image denoising using wavelet transforms , author=. Signal Processing: Image Communication , volume=

  228. [236]

    Performance of the maximum likelihood estimators for the parameters of multivariate generalized

    Bombrun, Lionel and Pascal, Fr. Performance of the maximum likelihood estimators for the parameters of multivariate generalized. 2012 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) , pages=. 2012 , organization=

    2012

  229. [237]

    Biometrics , volume=

    Multivariate elliptically contoured distributions for repeated measurements , author=. Biometrics , volume=

  230. [238]

    Communications in Statistics -- Theory and Methods , volume=

    A multivariate generalization of the power exponential family of distributions , author=. Communications in Statistics -- Theory and Methods , volume=

  231. [239]

    Hyperbolic distributions and distributions on hyperbolae , Year =

    Barndorff-Nielsen, O , Journal =. Hyperbolic distributions and distributions on hyperbolae , Year =

  232. [240]

    Journal of the American Statistical Association , volume =

    Hjort, Nils Lid and Claeskens, Gerda , title =. Journal of the American Statistical Association , volume =

  233. [241]

    Setting up alternating least squares and iterative majorization algorithms for solving various matrix optimization problems , Volume =

    Kiers, Henk A L , Journal =. Setting up alternating least squares and iterative majorization algorithms for solving various matrix optimization problems , Volume =

  234. [242]

    Browne and Paul D

    Ryan P. Browne and Paul D. McNicholas , Note =

  235. [243]

    Multivariate scale mixture of

    Eltoft, Torbj. Multivariate scale mixture of. Independent Component Analysis and Blind Signal Separation , Pages =

  236. [244]

    A new family of multivariate heavy-tailed distributions with variable marginal amounts of tailweights:

    Forbes, Florence and Wraith, Darren , Journal =. A new family of multivariate heavy-tailed distributions with variable marginal amounts of tailweights:. 2014 , volume =

    2014

  237. [245]

    Johnson, N. L. and Kotz, S. and Balakrishnan, N. , Publisher =. Continuous Univariate Distributions , Year =

  238. [246]

    , journal =

    Madigan, David and Raftery, Adrian E. , journal =. Model Selection and Accounting for Model Uncertainty in Graphical Models Using

  239. [247]

    Hoeting and Adrian E

    Jennifer A. Hoeting and Adrian E. Raftery and David Madigan , title =. 1999 , institution =

    1999

  240. [248]

    and Madigan, David and Raftery, Adrian E

    Hoeting, Jennifer A. and Madigan, David and Raftery, Adrian E. and Volinsky, Chris T. , journal =. Bayesian Model Averaging: A Tutorial , volume =

  241. [249]

    and McNicholas, P

    Wei, Y. and McNicholas, P. D. , title =. Advances in Data Analysis and Classification , year =

  242. [250]

    , Journal =

    Baricz, A. , Journal =. Tur‡n type inequalities for some probability density functions , Volume =

  243. [251]

    and Kozubowski, T

    Kotz, S. and Kozubowski, T. J. and Podgorski, K. , title =. 2001 , address =

    2001

  244. [252]

    McNeil, A. J. and Frey, R. and Embrechts, P. , title =. 2005 , location =

    2005

  245. [253]

    , year =

    Barndorff-Nielsen, O. , year =. Hyperbolic distributions and distributions on hyperbolae , journal =

  246. [254]

    Fraley and A

    C. Fraley and A. E. Raftery and T B Murphy and L Scrucca. mclust Version 4 for R : Normal Mixture Modeling for Model-Based Clustering, Classification, and Density Estimation. 2012

    2012

  247. [255]

    McNicholas, P. D. and Browne, R. P. , year = 2013, title =. Journal of the Royal Statistical Society: Series C , volume =

    2013

  248. [256]

    1978 , author =

    Estimating the dimension of a model , journal =. 1978 , author =

    1978

  249. [257]

    , year =

    Steinley, D. , year =. Properties of the. Psychological Methods , volume =

  250. [258]

    Browne, R. P. and McNicholas, P. D. and Sparling, M. D. , Title =. IEEE Transactions on Pattern Analysis and Machine Intelligence , Year =

  251. [259]

    Browne, R. P. and McNicholas, P. D. , Title =. Journal of Statistical Planning and Inference , Year =

  252. [260]

    Cattell, R. B. , year =. Personality and Motivation Structure and Measurement , publisher =

  253. [261]

    Tryon, R. C. , year =. Cluster Analysis , subtitle =

  254. [262]

    Tiedeman, D. V. , year =. On the Study of Types , editor =. Symposium on Pattern Analysis , address =

  255. [263]

    McNicholas, S. M. and McNicholas, P. D. and Browne, R. P. , Note =

  256. [264]

    and McNicholas, P

    Vrbik, I. and McNicholas, P. D. , Journal =. Parsimonious skew mixture models for model-based clustering and classification , Year =

  257. [265]

    Franczak, B. C. and Browne, R. P. and McNicholas, P. D. , Journal =. Mixtures of Shifted Asymmetric. 2014 , Volume =

    2014

  258. [266]

    Murray, P. M. and McNicholas, P. D. and Browne, R. B. , Journal =. A mixture of common skew- t factor analyzers , Year =

  259. [267]

    Murray, P. M. and Browne, R. B. and McNicholas, P. D. , Year =. Mixtures of skew-t factor analyzers , journal =

  260. [268]

    Steane, M. A. and McNicholas, P. D. and Yada, R. , Journal =. Model-based classification via mixtures of multivariate t-factor analyzers , Volume =

  261. [269]

    and McNicholas, P

    Lin, T.-I. and McNicholas, P. D. and Hsiu, J. H. , year =. Capturing patterns via parsimonious t mixture models , journal =

  262. [270]

    2013 , journal=

    On mixtures of skew normal and skew t-distributions , author=. 2013 , journal=

    2013

  263. [271]

    Statistical Methods and Applications , volume=

    Model-based clustering and classification with non-normal mixture distributions , author=. Statistical Methods and Applications , volume=. 2013 , publisher=

    2013

  264. [272]

    McLachlan and Sharon X

    Tsung-I Lin and Geoffrey J. McLachlan and Sharon X. Lee , Title =. Journal of Multivariate Analysis , Pages =. 2016 , Volume =

    2016

  265. [273]

    Murray, P. M. and Browne, R. B. and McNicholas, P. D. , Note =. Mixtures of `Unrestricted' Skew-t Factor Analyzers , Year =

  266. [274]

    Browne, R. P. and McNicholas, P. D. A mixture of generalized hyperbolic distributions. Canadian Journal of Statistics , volume=. 2015

    2015

  267. [275]

    2006 , journal =

    Aas, Kjersti and Haff, Ingrid Hob¾k , title =. 2006 , journal =

    2006

  268. [276]

    Scallan, A. J. , journal =. Maximum Likelihood Estimation for a Normal/

  269. [277]

    Shi and Selesnick, I

    F. Shi and Selesnick, I. W. , booktitle=. Multivariate Quasi-. 2006 , pages=

    2006

  270. [278]

    and Davison, AC and Goldstein, D.R

    Bhowmick, D. and Davison, AC and Goldstein, D.R. and Ruffieux, Y. , journal=. A. 2006 , publisher=

    2006

  271. [279]

    and Stathaki, T

    Mitianoudis, N. and Stathaki, T. , journal=. Overcomplete source separation using. 2005 , volume=

    2005

  272. [280]

    Cord and C

    A. Cord and C. Ambroise and J.-P. Cocquerez. Feature selection in robust clustering based on L aplace mixture. Pattern Recognition Letters. 2006

    2006

  273. [281]

    and Kim, T

    Eltoft, T. and Kim, T. and Lee, T. , Journal =. On the Multivariate

  274. [282]

    and Kozubowski, T

    Kotz, S. and Kozubowski, T. J. and Podgorski, K. , Edition =. The Laplace Distribution and Generalizations: A Revisit with Applications to Communications, Economics, Engineering, and Finance , Year =

  275. [283]

    Statistical Properties of the Generalized Inverse Gaussian Distribution , Year =

    J. Statistical Properties of the Generalized Inverse Gaussian Distribution , Year =

  276. [284]

    , Journal =

    Halgreen, C. , Journal =. Self-decomposibility of the generalized inverse

  277. [285]

    The shape of the generalized inverse

    Bl. The shape of the generalized inverse

  278. [286]

    , title =

    Bolin, D. , title =. 2011 , note =

    2011

  279. [287]

    Barndorff-Nielsen, O. E. , title =. Scandinavian Journal of Statistics , volume =

  280. [288]

    and Halgreen, C

    Barndorff-Nielsen, O. and Halgreen, C. , Journal =. Infinite divisibility of the hyperbolic and generalized inverse

  281. [289]

    Vrbik and P

    I. Vrbik and P. D. McNicholas , Journal =. Analytic calculations for the

  282. [290]

    , Journal =

    Lin, T.-I. , Journal =. Maximum likelihood estimation for multivariate skew normal mixture models , Volume =

  283. [291]

    Lee and G

    S. Lee and G. J. McLachlan , Note =. On the fitting of mixtures of multivariate skew t -distributions via the

  284. [292]

    and McLachlan, G

    Baek, J. and McLachlan, G. J. , Journal =. Mixtures of common t-factor analyzers for clustering high-dimensional microarray data , Volume =

  285. [293]

    McNicholas, P. D. and Subedi, S. , year =. Clustering gene expression time course data using mixtures of multivariate t-distributions , journal =

  286. [294]

    Edwards, A. W. F. and Cavalli-Sforza, L. L. , year =. A method for cluster analysis , journal =

  287. [295]

    Scott, A. J. and Symons, M. J. , year =. Clustering methods based on likelihood ratio criteria , journal =

  288. [296]

    Friedman, H. P. and Rubin, J. , year =. On some invariant criteria for grouping data , journal =

  289. [297]

    Discussion of `

    Gower, J C , year =. Discussion of `. Journal of the Royal Statistical Society: Series A , volume =

  290. [298]

    Proceedings of the Linnaean Society of New South Wales , volume =

    Johnson, L A S , title =. Proceedings of the Linnaean Society of New South Wales , volume =

  291. [299]

    1968 , note =

    Johnson, S C , title =. 1968 , note =

    1968

  292. [300]

    1970 , journal =

    Johnson, L A S , title =. 1970 , journal =

    1970

Pith tools

Reviewed August 2, 2026 · model on record in the stance chip above.