Pith. sign in

REVIEW 4 major objections 6 minor 55 references

Co-clustering of Response and Covariate Variables by Tri-Factorizing Their Non-negative Regression Coefficient Matrix

T0 review · 4 major / 6 minor · reviewed 2026-08-01 · deepseek-v4-flash

Pith's one-line read A non-negative tri-factorization of the regression coefficient matrix yields soft clusters on both sides of a two-block regression, with significance-tested links between them.

desk verdict A novel tri-factorized non-negative RRR with conditional inference; worth refereeing, but the abstract oversells the 'conservative after re-estimation' claim and per-path tests aren't factorization-invariant in the recommended Q<R regime. read the letter →

arxiv 2607.27474 v1 pith:M6KBUVE3 submitted 2026-07-29 stat.ME

classification stat.ME MSC 62H3062J0562H2562F03
keywords Co-clusteringNon-negativematrixtri-factorizationReduced-rankregressionBlock-correspondenceinferenceMultivariateTwo-blockdataConditionalWaldtestNMF-RRR
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 is trying to establish that the relationship between two blocks of variables measured on the same individuals can be read as a co-clustering of the variables themselves. It constrains the multivariate regression Y1 ≈ M Y2 so that the coefficient matrix factors as M = X1 Θ X2, where X1 softly groups response variables, X2 softly groups covariates, and the middle factor Θ records which covariate groups drive which response groups. The paper develops multiplicative updates for fitting, cross-validation for the two group counts, and a conditional Wald test for the entries of Θ. If correct, the method supplies interpretable, tested group-to-group links in settings such as microbiome–metabolome studies, where classical reduced-rank regression and canonical correlation analysis return signed factors or become ill-posed.

What carries the argument

The load-bearing object is the tri-factorization M = X1 Θ X2 of the regression coefficient matrix, with X1 and X2 non-negative and normalized so that their columns and rows sum to one, turning them into soft cluster profiles, and with Θ a non-negative Q×R parameter matrix. The middle factor Θ is what the paper identifies as the minimal structure that decouples the response and covariate groupings and carries the tested block correspondences. The machinery also includes multiplicative updates that preserve non-negativity and decrease the squared error, a normalization identity that ties the sum of the entries of Θ to the grand total of the coefficient matrix, and identifiability conditions (a

What would settle it

Run the full pipeline—rank selection by cross-validation, basis re-estimation, conditional Wald test—on data generated from the paper's identifiable Q=R=3 null design, but count every replication, including those where cross-validation fails to recover the true ranks. The paper reports sizes only after conditioning on correct rank recovery; an unconditional rejection rate above 5% would show that rank selection adds over-rejection that the conditioning hides, while a rate at or below nominal would support the conservative-size claim.

Watch

Extended reading notes

Core claim

The central claim is that the low-rank regression coefficient M = X1 Θ X2, with non-negative and normalized factors, is the natural algebraic object for two-block co-clustering: X1 and X2 are soft cluster profiles of response and covariate variables, and Θ is a Q×R block-correspondence matrix whose entries are estimated and tested. The estimator is called tri-factorized non-negative reduced-rank regression (NMF-RRR); because rank(M) ≤ min(Q,R), it belongs to the same model class as reduced-rank regression, but expresses the shared subspace in non-negative parts rather than signed directions. The paper establishes that fit is capped by min(Q,R), that a square Θ tends toward a near-permutation

Load-bearing premise

The load-bearing premise is that a significant entry of Θ can be read as a tested group-to-group link even though the test is conditional on bases and ranks estimated from the same data and, when Q < R, the covariate-side split is not uniquely identified.

Editorial extensions

If this is right

  • On microbiome–metabolome data, the method can recover a pronounced cross-structure in which each metabolite module is associated with two microbial groups, all paths significant, a pattern a square permutation Θ cannot express.
  • Because fit is capped by min(Q,R), the paper recommends taking Q ≤ R: extra response factors beyond the rank are not separately identifiable, while extra covariate factors can expose cross-structure.
  • When the number of covariates exceeds the sample size, the non-negative, normalized, low-rank parameterization remains well-behaved where unregularized CCA and RRR are ill-posed, at the price of lower in-sample fit.
  • Re-estimating the bases from the same data keeps the existence test conservative but makes the magnitudes of non-zero paths unreliable, so significant Θ entries should be read as tested presence or absence, not as effect sizes.
  • When covariate blocks are uncorrelated, unsupervised tri-NMF of the association matrix recovers the same clustering; the supervised factorization mainly buys accuracy when cross-block covariate correlation distorts the association.
  • Editorial extension: the paper leaves implicit that a formal selective-inference correction, conditioning on the full selection event rather than on the chosen bases, could convert the exploratory screen into confirmatory inference; until that exists, the method is best used for hypothesis generation.
  • Editorial extension: the one-fifth-of-dominant-path detection threshold observed in simulation suggests that a minimum-effect-size guideline could be pre-registered in future two-block studies, since weak paths are detectable only when dominant paths give up coefficient mass.
  • Editorial extension: because negative associations can only be encoded through membership in a different non-negative group, a compositional or count-data variant with a log-ratio or Poisson objective would plausibly change which groups emerge; this is a direct consequence of the modelling choice rather than of the factorization itself.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The paper proposes a tri-factorized non-negative reduced-rank regression (NMF-RRR) for two-block data, modeling Y1 ≈ X1 Θ X2 Y2 so that the non-negative regression coefficient matrix M = X1ΘX2 is a soft tri-factorization: X1 clusters response variables, X2 clusters covariate variables, and Θ is a block-correspondence matrix whose entries are estimated and tested. The paper derives multiplicative updates, rank selection by element-wise and sample-wise cross-validation, a conditional Wald test with sandwich standard errors, and identifiability results in Appendix A. It reports simulations on calibration and four applications (Doubs, nutrimouse, FRANZOSA, Wine). The manuscript is unusually candid about limitations: Section 5 states that for Q<R the null θqr=0 is defined only relative to the selected factorization, and Section 7 restricts per-path type-I error simulations to identifiable Q=R designs conditional on rank recovery.

Significance. If the claims were fully supported, the method would fill a genuine gap: a two-sided soft co-clustering of response and covariate variables with an estimable block-correspondence matrix, connecting reduced-rank regression and non-negative matrix factorization. Strengths include the explicit identifiability analysis (Appendix A), transparent simulation reporting, reproducible code and data scripts, and the honest disclosure of the conditional nature of the inference. However, the paper's central inferential selling point—a 'tested matrix of block correspondences'—is narrower than the abstract suggests. The 'conservative after re-estimation' claim is established only for identifiable Q=R designs conditional on rank recovery (Table 10), while the recommended Q<R regime lacks a decomposition-invariant per-path null. The method is still valuable as an exploratory co-clustering and estimation tool, but the path tests in Q<R applications should be described as conditional screens, not confirmatory tests.

major comments (4)
  1. [§4.3, §5, §7, Appendix A Prop. 5(c)] The abstract's claim that the conditional Wald test is 'conservative after re-estimation' is not supported in the regime the paper recommends. Section 4.3 recommends Q≤R and uses Q<R to expose cross-structure (nutrimouse Q=2,R=3; FRANZOSA Q=2,R=4), but Section 5 and Appendix A Prop. 5(c) state that for Q<R the split into Θ and X2 is not uniquely identified, so the null θqr=0 is defined only relative to the selected factorization. The simulations in Section 7 restrict per-path size to identifiable Q=R designs, and Table 10 reports sizes conditional on recovering the true ranks (74.6% recovery in the synthetic design). The abstract should be revised to say 'conservative in the identifiable Q=R designs examined, conditional on rank recovery,' and the paper should either provide Q<R simulations that evaluate path tests under the selected factorization or explicitly frame all Q<R path tests a
  2. [§4.1] The masked (weighted) updates for element-wise cross-validation are delegated to the nmfkc package without derivation. Since the rank-selection rule (12) and the full-pipeline simulation (Table 10) depend on these updates, the paper should provide the explicit update formulas and state whether monotonicity/KKT properties analogous to Proposition 4 hold. As written, a central algorithmic component is a black box, which hampers reproducibility and verification.
  3. [§7, Tables 9–10] The 'existence test stays safe' conclusion is based on very limited evidence. Table 9 shows empirical size 0.000 at only two true-zero paths in a high-signal Doubs-calibrated design; this is extreme conservatism, not a demonstration of controlled size. Table 10 conditions on recovering the true ranks and excludes 25.4% of replications in the synthetic design. The unconditional behavior of the full procedure is not reported, and the paper itself notes that an unconditional per-path type-I error is not well defined when Q<R. This limitation should be stated in the abstract and in the conclusions, not only in the simulation section, because the real-data results in Section 6 report significant paths from Q<R analyses.
  4. [§5, §6, Table 11] The wild bootstrap is presented as providing interval estimates, but its calibration is simulated only with fixed bases and correlated errors (Table 11), not with re-estimated bases or rank selection. Given that Table 9 shows severe bias in path magnitudes after basis re-estimation, the bootstrap intervals used in the real-data analyses should either be accompanied by a re-estimation simulation or be labeled as exploratory. Currently the text in Section 6 describes significant Θ entries as 'tested present/absent links' while magnitudes are 'conditional on the estimated bases,' but the absence of bootstrap calibration under re-estimation leaves the interval claims unverified.
minor comments (6)
  1. [Abstract] The phrase 'conservative after re-estimation' should be qualified as 'conservative in the identifiable Q=R designs examined, conditional on rank recovery.' As written, it overstates the scope of the simulation evidence.
  2. [§7, Table 9] Calling empirical size 0.000 'conservative' is correct but misleading; the test never rejects at the null paths, which is far more conservative than nominal. Consider describing it as 'over-conservative in these settings' and noting that this may reflect low sensitivity rather than ideal control.
  3. [§8] The statement that re-fitting the four analyses under the KL objective 'left the significant structure intact' is important but is not accompanied by a table or numeric summary. Please provide the results in an appendix or online supplement.
  4. [§6.1] The reference to 'Huet's classical longitudinal zonation' is not in the reference list. Please add a citation.
  5. [§5, Eq. (14)] The notation 'σ2IP1N' should be typeset as σ^2 I_{P1N} for consistency with the rest of the paper.
  6. [§4.1] Equation (12) uses σ(Q,R) for RMSE; later in Section 7 the one-standard-error rule is applied to σ². Clarify this distinction in the text.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: NMF-RRR is a constrained reparameterization of rank-min(Q,R) regression; the Θ inference is a disclosed conditional Wald test, and the paper's own simulations provide independent calibration evidence.

full rationale

The central construction M=X1ΘX2 is defined independently of the target claims. Proposition 1 follows from the rank inequality rank(M)≤min(Q,R) and the definition of rank-r RRR; there is no circular identity in which a fitted quantity is reused as a prediction. The multiplicative updates (8)-(10) are derived from the gradient of the loss (4), and Propositions 3-4 are standard convexity/multiplicative-update statements. Rank selection uses element-wise or sample-wise cross-validation on held-out entries, not the training fit. The inference for Θ is explicitly conditional on the estimated bases: Section 5 states 'not the formal selective-inference construction... but a Wald test conditional on the chosen bases' and acknowledges that for Q<R 'the null θqr=0 is defined only relative to the selected factorization, so "presence of a path" is not a decomposition-invariant hypothesis.' This is a disclosed post-selection/identifiability limitation, not a definitional equivalence. Table 10 conditions on correct rank recovery and the paper says so; unconditional behavior is not claimed. The simulations (Tables 8-12) are self-contained numerical evidence, and the data analyses use external public benchmarks. Self-citations to Satoh (2025, 2026a-c) provide prior NMF-with-covariates context and an initialization scheme, but the central derivation, updates, simulations, and identifiability appendix (Proposition 5, with proof) do not rest on those citations. No fit parameter is renamed as a prediction; no uniqueness theorem is imported from the authors' prior work; and the paper explicitly disclaims that the tri-factorization is a new matrix class ('not more general than a rank-r bi-factorization; it is a reparameterization'). Accordingly, no circular step is exhibited.

Assumptions & free parameters 3 free parameters · 4 assumptions · 1 invented entities

The central claim rests on a specific working model for inference (Gaussian errors, independence across individuals), on the identifiability assumption of anchor variables, and on analyst-chosen ranks and screening thresholds. The method introduces no new physical entities; the latent clusters are fitted constructs without independent evidence.

free parameters (3)
  • Ranks (Q,R) = Doubs (2,2), nutrimouse (2,3), FRANZOSA (2,4), Wine (3,3)
    Chosen by element-wise cross-validation with a one-standard-error rule; the co-clustering and all subsequent inference depend on these choices, and the paper conditions on them (Sections 4.1, 6).
  • FRANZOSA screening size = 30 variables per block (robustness at 50/100/200/400)
    Two blocks reduced from 11,720 and 8,848 features to the 30 most variable high-prevalence variables; group memberships are threshold-dependent (Section 6.3).
  • Min–max transform = per-variable [0,1] mapping
    Modeling choice making signed variables non-negative; encodes negative associations indirectly via group membership and distorts zero structure (Sections 2.1, 8).
assumptions (4)
  • domain assumption Both blocks are non-negative after transform; the working model for inference is Gaussian, vec(E)~N(0,σ²I_{P1N}) (eq. 13)
    Section 5 introduces the Gaussian working model solely for inference; the paper acknowledges it is a working assumption and that the sandwich SEs are meant to cover its misspecification.
  • domain assumption Independence across individuals in the sandwich variance and wild bootstrap (eq. 15)
    Section 5: 'summing the score outer products over individuals treats individuals as independent'.
  • domain assumption Separability (anchor-variable) condition for identifiability of X1 and G=ΘX2 (Appendix A)
    Prop. 5(a) needs each response group to own a variable loading on it alone; the paper does not verify this on any data set, and for Q<R the split of ΘX2 into Θ and X2 remains non-identified.
  • domain assumption Y2 full row rank for identification of the coefficient M (for the fixed-total interpretation of Prop. 2)
    The paper notes in Section 3.5 that when P2>N (nutrimouse), only MY2 is identified.
invented entities (1)
  • Soft clusters (latent groups) of response and covariate variables
    purpose: Interpretive units of NMF-RRR: the co-clustering output; formed by normalized non-negative basis columns/rows.
    The clusters are fitted constructs with no external ground truth; the paper validates them only by agreement with unsupervised tri-NMF (Doubs ARI 1.00) and biological plausibility, which is circumstantial.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Co-clustering of Response and Covariate Variables by Tri-Factorizing Their Non-negative Regression Coefficient Matrix." pith.science (2026). https://pith.science/paper/M6KBUVE3

@misc{pith2026260727474,
  author       = {Pith},
  title        = {Pith review of: Co-clustering of Response and Covariate Variables by Tri-Factorizing Their Non-negative Regression Coefficient Matrix},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/M6KBUVE3}},
  note         = {Machine review of arXiv:2607.27474}
}
abstract

Two-block data---two sets of variables measured on the same individuals, such as microbial taxa and metabolites---raise the question of how \emph{groups} of covariate variables relate to \emph{groups} of response variables. Co-clustering answers this for a single matrix, not for two variable blocks; existing two-block methods either cluster only one side or return signed factors rather than clusters. Starting from the multivariate linear regression $Y_1\approx M Y_2$, we give its non-negative coefficient matrix a tri-factorization $M=X_1\Theta X_2$ (a tri-NMF), so that $X_1$ softly clusters the response variables, $X_2$ the covariate variables, and $\Theta$ is a tested matrix of block correspondences. This makes the method the non-negative member of the reduced-rank regression (RRR) family, expressing RRR's low-rank class in a parts-based basis as NMF relates to PCA; the constraint can only restrict the fit, so predictive accuracy is not the aim; the co-clustering and tested correspondences are. We give multiplicative update rules, choose the two ranks by cross-validation, and develop a conditional Wald test for $\Theta$ applied after basis selection; its size is nominal with fixed bases, conservative after re-estimation, and slightly above nominal under correlated responses, while a non-zero path's \emph{magnitude} stays conditional on the estimated bases. We illustrate the method---a tri-factorized non-negative RRR (NMF-RRR)---on four data sets spanning a permutation structure (Doubs, community ecology), a weak cross-structure under $p>n$ (nutrimouse, nutrigenomics), a pronounced one in a screened microbiome--metabolome study (FRANZOSA, where two microbial groups are jointly associated with each metabolite module), and a classification special case (Wine).

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

55 extracted references · 20 canonical work pages

  1. [1]

    UCI Machine Learning Repository, DOI : 10.24432/C5PC7J

    Aeberhard S, Forina M (1992) Wine. UCI Machine Learning Repository, DOI : 10.24432/C5PC7J

  2. [2]

    Annals of Mathematical Statistics 22(3):327--351

    Anderson TW (1951) Estimating linear restrictions on regression coefficients for multivariate normal distributions. Annals of Mathematical Statistics 22(3):327--351. DOI :10.1214/aoms/1177729580

  3. [3]

    In: Proceedings of the 44th Annual ACM Symposium on Theory of Computing (STOC), pp 145--162, DOI :10.1145/2213977.2213994

    Arora S, Ge R, Kannan R, et al (2012) Computing a nonnegative matrix factorization---provably. In: Proceedings of the 44th Annual ACM Symposium on Theory of Computing (STOC), pp 145--162, DOI :10.1145/2213977.2213994

  4. [4]

    Analytical and Bioanalytical Chemistry 390(5):1241--1251

    Bro R, Kjeldahl K, Smilde AK, et al (2008) Cross-validation of component models: A critical look at current methods. Analytical and Bioanalytical Chemistry 390(5):1241--1251. DOI :10.1007/s00216-007-1790-1

  5. [5]

    Annals of Statistics 40(5):2359--2388

    Bunea F, She Y, Wegkamp MH (2012) Joint variable and rank selection for parsimonious estimation of high-dimensional matrices. Annals of Statistics 40(5):2359--2388. DOI :10.1214/12-AOS1039

  6. [6]

    IEEE Transactions on Pattern Analysis and Machine Intelligence 33(8):1548--1560

    Cai D, He X, Han J, et al (2011) Graph regularized nonnegative matrix factorization for data representation. IEEE Transactions on Pattern Analysis and Machine Intelligence 33(8):1548--1560. DOI :10.1109/TPAMI.2010.231

  7. [7]

    Journal of the Royal Statistical Society: Series B 74(2):203--221

    Chen K, Chan KS, Stenseth NC (2012) Reduced rank stochastic regression with a sparse singular value decomposition. Journal of the Royal Statistical Society: Series B 74(2):203--221. DOI :10.1111/j.1467-9868.2011.01002.x

  8. [8]

    Journal of the American Statistical Association 107(500):1533--1545

    Chen L, Huang JZ (2012) Sparse reduced-rank regression for simultaneous dimension reduction and variable selection. Journal of the American Statistical Association 107(500):1533--1545. DOI :10.1080/01621459.2012.734178

Show all 55 references
  1. [9]

    In: Proceedings of the 8th International Conference on Intelligent Systems for Molecular Biology (ISMB), pp 93--103

    Cheng Y, Church GM (2000) Biclustering of expression data. In: Proceedings of the 8th International Conference on Intelligent Systems for Molecular Biology (ISMB), pp 93--103

  2. [10]

    In: Proceedings of the 7th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD), pp 269--274, DOI :10.1145/502512.502550

    Dhillon IS (2001) Co-clustering documents and words using bipartite spectral graph partitioning. In: Proceedings of the 7th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD), pp 269--274, DOI :10.1145/502512.502550

  3. [11]

    Journal of Machine Learning Research 3:1265--1287

    Dhillon IS, Mallela S, Kumar R (2003 a ) A divisive information-theoretic feature clustering algorithm for text classification. Journal of Machine Learning Research 3:1265--1287

  4. [12]

    In: Proceedings of the 9th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD), pp 89--98, DOI :10.1145/956750.956764

    Dhillon IS, Mallela S, Modha DS (2003 b ) Information-theoretic co-clustering. In: Proceedings of the 9th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD), pp 89--98, DOI :10.1145/956750.956764

  5. [13]

    In: Proceedings of the SIAM International Conference on Data Mining (SDM), pp 606--610, DOI :10.1137/1.9781611972757.70

    Ding C, He X, Simon HD (2005) On the equivalence of nonnegative matrix factorization and spectral clustering. In: Proceedings of the SIAM International Conference on Data Mining (SDM), pp 606--610, DOI :10.1137/1.9781611972757.70

  6. [14]

    In: Proceedings of the 12th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD), pp 126--135, DOI :10.1145/1150402.1150420

    Ding C, Li T, Peng W, et al (2006) Orthogonal nonnegative matrix tri-factorizations for clustering. In: Proceedings of the 12th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD), pp 126--135, DOI :10.1145/1150402.1150420

  7. [15]

    Donoho D, Stodden V (2003) When does non-negative matrix factorization give a correct decomposition into parts? In: Advances in Neural Information Processing Systems (NeurIPS)

  8. [16]

    Psychometrika 1(3):211--218

    Eckart C, Young G (1936) The approximation of one matrix by another of lower rank. Psychometrika 1(3):211--218. DOI :10.1007/BF02288367

  9. [17]

    BMC Bioinformatics 23:529

    Fitzgerald T, Jones A, Engelhardt BE (2022) A P oisson reduced-rank regression model for association mapping in sequencing data. BMC Bioinformatics 23:529. DOI:10.1186/s12859-022-05054-6

  10. [18]

    PLoS Pathogens 15(3):e1007581

    Foley MH, O'Flaherty S, Barrangou R, et al (2019) Bile salt hydrolases: gatekeepers of bile acid metabolism and host--microbiome crosstalk in the gastrointestinal tract. PLoS Pathogens 15(3):e1007581. DOI:10.1371/journal.ppat.1007581

  11. [19]

    Nature Microbiology 4:293--305

    Franzosa EA, Sirota-Madi A, Avila-Pacheco J, et al (2019) Gut microbiome structure and metabolic activity in inflammatory bowel disease. Nature Microbiology 4:293--305. DOI:10.1038/s41564-018-0306-4

  12. [20]

    IEEE Signal Processing Magazine 36(2):59--80

    Fu X, Huang K, Sidiropoulos ND, et al (2019) Nonnegative matrix factorization for signal and data analytics: identifiability, algorithms, and applications. IEEE Signal Processing Magazine 36(2):59--80. DOI :10.1109/MSP.2018.2877582

  13. [21]

    Wiley-ISTE, DOI :10.1002/9781118649480

    Govaert G, Nadif M (2013) Co-Clustering: Models, Algorithms and Applications. Wiley-ISTE, DOI :10.1002/9781118649480

  14. [22]

    Biometrika 28(3/4):321--377

    Hotelling H (1936) Relations between two sets of variates. Biometrika 28(3/4):321--377. DOI :10.1093/biomet/28.3-4.321

  15. [23]

    Journal of Multivariate Analysis 5(2):248--264

    Izenman AJ (1975) Reduced-rank regression for the multivariate linear model. Journal of Multivariate Analysis 5(2):248--264. DOI :10.1016/0047-259X(75)90042-1

  16. [24]

    Computational Intelligence and Neuroscience 2008:764206

    Laurberg H, Christensen MG, Plumbley MD, et al (2008) Theorems on positive data: on the uniqueness of NMF . Computational Intelligence and Neuroscience 2008:764206. DOI :10.1155/2008/764206

  17. [25]

    Nature 401(6755):788--791

    Lee DD, Seung HS (1999) Learning the parts of objects by non-negative matrix factorization. Nature 401(6755):788--791. DOI :10.1038/44565

  18. [26]

    In: Advances in Neural Information Processing Systems (NeurIPS), pp 556--562

    Lee DD, Seung HS (2000) Algorithms for non-negative matrix factorization. In: Advances in Neural Information Processing Systems (NeurIPS), pp 556--562

  19. [27]

    IEEE Signal Processing Letters 17(1):4--7

    Lee H, Yoo J, Choi S (2010) Semi-supervised nonnegative matrix factorization. IEEE Signal Processing Letters 17(1):4--7. DOI :10.1109/LSP.2009.2027163

  20. [28]

    IEEE Transactions on Neural Networks 18(6):1589--1596

    Lin CJ (2007) On the convergence of multiplicative update algorithms for nonnegative matrix factorization. IEEE Transactions on Neural Networks 18(6):1589--1596

  21. [29]

    Annals of Statistics 16(4):1696--1708

    Liu RY (1988) Bootstrap procedures under some non-i.i.d.\ models. Annals of Statistics 16(4):1696--1708. DOI :10.1214/aos/1176351062

  22. [30]

    Annals of Statistics 21(1):255--285

    Mammen E (1993) Bootstrap and wild bootstrap for high dimensional linear models. Annals of Statistics 21(1):255--285. DOI :10.1214/aos/1176349025

  23. [31]

    Hepatology 45(3):767--777

    Martin PGP, Guillou H, Lasserre F, et al (2007) Novel aspects of PPAR -mediated regulation of lipid and xenobiotic metabolism revealed through a nutrigenomic study. Hepatology 45(3):767--777. DOI: 10.1002/hep.21510

  24. [32]

    Journal of Computational and Graphical Statistics 26(4):814--825

    Mishra A, Dey DK, Chen K (2017) Sequential co-sparse factor regression. Journal of Computational and Graphical Statistics 26(4):814--825. DOI: 10.1080/10618600.2017.1340891

  25. [33]

    Annals of Applied Statistics 3(2):564--594

    Owen AB, Perry PO (2009) Bi-cross-validation of the svd and the nonnegative matrix factorization. Annals of Applied Statistics 3(2):564--594. DOI :10.1214/08-AOAS227

  26. [34]

    Biometrika 51(3/4):313--326

    Potthoff RF, Roy SN (1964) A generalized multivariate analysis of variance model useful especially for growth curve problems. Biometrika 51(3/4):313--326. DOI :10.1093/biomet/51.3-4.313

  27. [35]

    Springer, DOI :10.1007/978-1-4757-2853-8

    Reinsel GC, Velu RP (1998) Multivariate Reduced-Rank Regression: Theory and Applications. Springer, DOI :10.1007/978-1-4757-2853-8

  28. [36]

    Current Opinion in Gastroenterology 30(3):332--338

    Ridlon JM, Kang DJ, Hylemon PB, et al (2014) Bile acids and the gut microbiome. Current Opinion in Gastroenterology 30(3):332--338. DOI:10.1097/MOG.0000000000000057

  29. [37]

    Japanese Journal of Applied Statistics 52(2):59--74

    Satoh K (2023) On non-negative matrix factorization using G aussian kernels as covariates (in J apanese). Japanese Journal of Applied Statistics 52(2):59--74. In Japanese. DOI:10.5023/jappstat.52.59

  30. [38]

    ArXiv:2403.05359

    Satoh K (2025) Applying non-negative matrix factorization with covariates to the longitudinal data as growth curve model. ArXiv:2403.05359

  31. [39]

    Japanese Journal of Statistics and Data Science DOI:10.1007/s42081-026-00349-x; arXiv:2510.10375

    Satoh K (2026 a ) Applying non-negative matrix factorization with covariates to label matrix for classification. Japanese Journal of Statistics and Data Science DOI:10.1007/s42081-026-00349-x; arXiv:2510.10375

  32. [40]

    Japanese Journal of Statistics and Data Science 9(1):79--97

    Satoh K (2026 b ) Applying non-negative matrix factorization with covariates to multivariate time series as a vector autoregression model. Japanese Journal of Statistics and Data Science 9(1):79--97. DOI:10.1007/s42081-025-00314-0; arXiv:2501.17446

  33. [41]

    doi:10.32614/CRAN.package.nmfkc, ://CRAN.R-project.org/package=nmfkc, r package version 0.8.8

    Satoh K (2026 c ) nmfkc: Non-Negative Matrix Factorization with Kernel Covariates. doi:10.32614/CRAN.package.nmfkc, ://CRAN.R-project.org/package=nmfkc, r package version 0.8.8

  34. [42]

    Trends in Biotechnology 33(9):496--503

    Shin NR, Whon TW, Bae JW (2015) Proteobacteria: microbial signature of dysbiosis in gut microbiota. Trends in Biotechnology 33(9):496--503. DOI:10.1016/j.tibtech.2015.06.011

  35. [43]

    Proceedings of the National Academy of Sciences 112(25):7629--7634

    Taylor J, Tibshirani RJ (2015) Statistical learning and selective inference. Proceedings of the National Academy of Sciences 112(25):7629--7634. DOI :10.1073/pnas.1507583112

  36. [44]

    In: Proceedings of the 37th Allerton Conference on Communication, Control and Computing, pp 368--377

    Tishby N, Pereira FC, Bialek W (1999) The information bottleneck method. In: Proceedings of the 37th Allerton Conference on Communication, Control and Computing, pp 368--377

  37. [45]

    AIP Advances 10(10):105110

    Tokuda Y, Fujisawa M, Packwood DM, et al (2020) Data-driven design of glasses with desirable optical properties using statistical regression. AIP Advances 10(10):105110. DOI:10.1063/5.0022451

  38. [46]

    AIP Advances 11(12):125127

    Tokuda Y, Fujisawa M, Ogawa J, et al (2021) A machine learning approach to the prediction of the dispersion property of oxide glass. AIP Advances 11(12):125127. DOI:10.1063/5.0075425

  39. [47]

    IEEE Transactions on Pattern Analysis and Machine Intelligence 39(3):417--429

    Trigeorgis G, Bousmalis K, Zafeiriou S, et al (2017) A deep matrix factorization method for learning attribute representations. IEEE Transactions on Pattern Analysis and Machine Intelligence 39(3):417--429. DOI :10.1109/TPAMI.2016.2554555

  40. [48]

    IEEE Transactions on Information Theory 65(8):4924--4939

    Uematsu Y, Fan Y, Chen K, et al (2019) SOFAR : Large-scale association network learning. IEEE Transactions on Information Theory 65(8):4924--4939. ArXiv:1704.08349

  41. [49]

    PhD thesis, Universit\'e de Franche-Comt\'e, Besan c on

    Verneaux J (1973) Cours d'eau de franche-comt\'e (massif du jura): recherches \'ecologiques sur le r\'eseau hydrographique du doubs; essai de biotypologie. PhD thesis, Universit\'e de Franche-Comt\'e, Besan c on

  42. [50]

    In: Gani J (ed) Perspectives in Probability and Statistics

    Wold H (1975) Soft modelling by latent variables: the non-linear iterative partial least squares (nipals) approach. In: Gani J (ed) Perspectives in Probability and Statistics. Academic Press, p 117--142, DOI :10.1017/S0021900200047604. Festschrift published as J.\ Appl.\ Proba...

  43. [51]

    Technometrics 20(4):397--405

    Wold S (1978) Cross-validatory estimation of the number of components in factor and principal components models. Technometrics 20(4):397--405. DOI :10.1080/00401706.1978.10489693

  44. [52]

    Annals of Statistics 14(4):1261--1295

    Wu CFJ (1986) Jackknife, bootstrap and other resampling methods in regression analysis. Annals of Statistics 14(4):1261--1295. DOI :10.1214/aos/1176350142

  45. [53]

    Frontiers in Big Data 2:27

    Yu M, Ramamurthy KN, Thompson A, et al (2019) Simultaneous parameter learning and bi-clustering for multi-response models. Frontiers in Big Data 2:27. DOI: 10.3389/fdata.2019.00027

  46. [54]

    IEEE Transactions on Neural Networks 17(3):683--695

    Zafeiriou S, Tefas A, Buciu I, et al (2006) Exploiting discriminant information in nonnegative matrix factorization with application to frontal face verification. IEEE Transactions on Neural Networks 17(3):683--695. DOI :10.1109/TNN.2006.873291

  47. [55]

    In: PRICAI 2006: Trends in Artificial Intelligence, pp 404--412, DOI :10.1007/978-3-540-36668-3\_44

    Zhang D, Zhou ZH, Chen S (2006) Non-negative matrix factorization on kernels. In: PRICAI 2006: Trends in Artificial Intelligence, pp 404--412, DOI :10.1007/978-3-540-36668-3\_44

Pith tools

Reviewed August 1, 2026 · model on record in the stance chip above.