Pith. sign in

REVIEW 4 major objections 4 minor 2 cited by

Robust High-Dimensional Covariate-Assisted Network Modeling

T0 review · 4 major / 4 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read A latent-space network model with horseshoe priors on covariate coefficients and node mismatches contracts at rate (1+s_b+s_x)d log(np)/n^2 and improves misclustering whenever a sparse covariate signal exists.

desk verdict The method is genuinely new and worth a serious look, but the advertised community-detection improvement rests on a dropped sqrt(n) in the proof of Theorem 3.3. read the letter →

arxiv 2505.02986 v1 pith:C4PFQOBE submitted 2025-05-05 stat.ME stat.CO

Peng Zhao , Yabo Niu This is my paper · ORCID
classification stat.MEstat.CO MSC 62F1562H30
keywords networkanalysiscommunitydetectionhigh-dimensionalcovariateslatentspacemodelhorseshoepriorfractionalposteriorvariationalinferencesparsenetworks
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper proposes a Bayesian latent-space model for networks with high-dimensional node covariates and claims that putting horseshoe shrinkage priors on two layers—the covariate-to-latent coefficient rows and the node-level discrepancies between network and covariate information—lets the posterior borrow covariate strength only where it is useful. The central theoretical claim is that the fractional posterior for link probabilities contracts at rate $\epsilon_n^2 = M_0^2(1+s_b+s_x) d \log(np)/n^2$ in average $\alpha$-Rényi divergence, and that $K$-means on the estimated latent vectors misclusters at rate of order $e^{-\beta^*}\epsilon_n^2/\delta^2$, improving the network-only rate $e^{-\beta^*}d \log n/(n\delta^2)$ when few covariates are informative and few nodes mismatch. If correct, this gives practitioners one procedure that improves network estimation and clustering when a sparse covariate signal exists and degrades gracefully to network-only behavior when covariates are useless or misleading. Two variational algorithms, coordinate-ascent and stochastic, are developed so the method scales to networks with tens of thousands of nodes.

What carries the argument

The load-bearing object is the discrepancy matrix $X-ZB$ between latent network positions and a low-rank linear transformation of the covariates. The model places independent horseshoe global-local shrinkage priors, built from half-Cauchy scales, on each row $b_j$ of $B$ and on each residual $x_i-B^\top z_i$, so the posterior can shrink irrelevant covariate rows to zero and, node by node, shrink the covariate-driven approximation toward the network-learned position only when they agree. The proof machinery is the fractional posterior analyzed through $\alpha$-Rényi divergence, a new $\ell_{2,1}$ prior-concentration lemma for multivariate shrinkage priors, and a Chernoff-type Gaussian concentration bound; the computation is made conjugate by the tangent transform for the Bernoulli likelihood and an inverse-Gamma mixture representation of the half-Cauchy prior. A restricted eigenvalue condition on $Z$ translates deviations in $B$ into deviations in $ZB$.

What would settle it

Generate networks from the model with true latent vectors that violate the boundedness-and-centering event, for example by letting their norms grow with $n$ or giving them a nonzero row mean, then fit the published CAVI/SVI algorithms and measure misclustering as $n$ grows; if the error does not track $e^{-\beta^*}(1+s_b+s_x)d\log(np)/(n^2\delta^2)$ but instead tracks the network-only $e^{-\beta^*}d\log n/(n\delta^2)$, the projection assumption is load-bearing.

Watch

Extended reading notes

Core claim

The paper's central claim is that simultaneous row-sparsity on the covariate loading matrix $B^*$ and on the discrepancy matrix $X^*-ZB^*$ makes covariate-assisted network modeling adaptive: the fractional posterior recovers connection probabilities at rate $\epsilon_n^2 = M_0^2(1+s_b+s_x)d\log(np)/n^2$, which beats the network-only rate $d\log n/n$ whenever $s_b+s_x=o(n)$ and $\log p=O(\log n)$. Under an additional projection prior that bounds and centers the latent vectors—stated to be used only in the proof, not in the algorithm—the inner products $x_i^\top x_j$ and the intercept $\beta$ contract with an extra $e^{-\beta^*}$ factor, and $K$-means on the posterior estimate of $X$ misclusters at rate $O(e^{-\beta^*}\epsilon_n^2/\delta^2)$, improving on the network-only rate $O(e^{-\beta^*}d\log n/(n\delta^2))$. When $B^*=0$, so covariates carry no signal, the method reverts to the network-only rate, and when covariates and latent positions agree perfectly ($s_x=0$), it recovers the shared-latent-vector model. The authors further show the rate holds without requiring the covariate signal to be strong enough for exact variable selection; only the prediction error $\|ZB-X\|_F$ needs to be small.

Load-bearing premise

The formal latent-vector and clustering guarantees hold for a projected prior that bounds and centers the latent vectors, but the published algorithms do not apply that projection; if the projection materially changes posterior behavior, the improved rates may not hold for the method as actually run.

Editorial extensions

If this is right

  • Link-probability estimation error drops below the network-only rate whenever $s_b+s_x=o(n)$ and $\log p=O(\log n)$, and this gain does not require strong-enough covariate signals for exact variable selection.
  • Community detection becomes consistent at cluster separation $\delta=\omega(e^{-\beta^*/2}\epsilon_n)$, which for constant $s_b,s_x$ lowers the required separation from roughly $1/\sqrt{n}$ to roughly $1/n$ in sparse regimes.
  • If the covariates are pure noise, the prior drives $B$ toward zero and the method contracts at the same rate as a network-only latent-space model, so full covariate misspecification does not break the procedure.
  • The coordinate-ascent algorithm costs $O(n^2+p^2)$ per iteration and the stochastic variant costs $O(|E|)$ per iteration, making the approach feasible for networks with tens of thousands of nodes and high-dimensional covariates.

Reading between the lines

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

  • The rate's dependence on $s_x$ suggests treating each mismatched node as costing one extra $d\log(np)/n^2$ term; an implication is that the improved rate survives only when the mismatch fraction is $o(n/\log p)$, giving a simple rule of thumb for when covariate information should be included.
  • Because the misclustering theorem is proved under the projected prior while the published algorithms do not enforce the projection, a direct empirical check would compare variational posterior draws against the bounded-and-centered event; if the draws frequently leave it, the formal rate should be re-derived for the unprojected family.
  • The same two-layer shrinkage structure could be reused for multilayer or dynamic networks where one covariate matrix informs several layers; the paper lists this as future work, and the rate analysis suggests the improvement would appear only if the shared signal is sparse in the same two-layer sense.
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 / 4 minor

Summary. This paper proposes a Bayesian latent space model (CALSM) for network data with high-dimensional node covariates. The model posits X approximately equal to ZB with B row-sparse, and uses horseshoe-type shrinkage on both the covariate coefficients and the node-level deviations from ZB. The authors develop two variational inference algorithms (CAVI with an R/Rcpp implementation, and SVI with edge subsampling for large sparse networks) and analyze the fractional posterior in the homogeneous sparse network setting. The main theoretical results are Theorem 3.1, giving contraction of the fractional posterior for the link-probability matrix in average alpha-Renyi divergence at rate epsilon_n^2 = M_0^2(1+s_b+s_x)d log(np)/n^2; Corollary 3.2, controlling the average squared error of inner products with an e^{-beta*} factor; and Theorem 3.3, which states a misclustering rate after K-means of L(Xihat, X*) less than or similar to e^{-beta*} epsilon_n^2/delta^2 and claims an improvement over the network-only rate e^{-beta*} d log n/(n delta^2). Simulations and three real-world datasets, plus a large PubMed example, are used to demonstrate estimation and clustering performance.

Significance. The paper addresses a relevant problem, and the general strategy of using shrinkage priors on both covariate coefficients and node-level discrepancies is natural. If the stated rates were correct, the model would deliver a principled way to adaptively combine network and covariate information, degrading gracefully when covariates are uninformative. The manuscript is transparent: it explicitly states that Assumption 6 (a projection to bounded and centered latent vectors) is used only for the proofs and is not used in the implemented algorithms, and it provides detailed appendices with prior-concentration bounds and algorithm derivations. The principal importance of the paper rests on the community-detection improvement claimed in Theorem 3.3, and that claim is not supported by the proof as written.

major comments (4)
  1. [Theorem 3.3 / Appendix A.4] The proof of Theorem 3.3 drops a factor of n when converting the inner-product bound of Corollary 3.2 into a bound on the latent positions. Corollary 3.2 controls (1/n^2) sum_{i,j} (x_i'x_j - x_i*'x_j*)^2 <= M e^{-beta*} epsilon_n^2, hence ||X X' - X* X*'||_F less than or similar to n e^{-beta*/2} epsilon_n. Lemma A.6 then yields ||X - X*O||_F less than or similar to sqrt(n) e^{-beta*/2} epsilon_n, not e^{-beta*/2} epsilon_n as written in Appendix A.4. Plugging this into Lemma A.7 gives L(Xihat, X*) less than or similar to n e^{-beta*} epsilon_n^2/delta^2, matching the block-size condition (4) but contradicting the displayed rate e^{-beta*} epsilon_n^2/delta^2 in Theorem 3.3. Since epsilon_n^2 = (1+s_b+s_x)d log(np)/n^2, the corrected misclustering rate is e^{-beta*}(1+s_b+s_x)d log(np)/(n delta^2), which is the same order as the network-only rate e^{-beta*} d log n/(n delta^2) when p is polynomial and s_b+s_x are bounded, and worse when these constants grow. The advertised adaptive improvement in community detection is therefore not established.
  2. [Section 3, Assumption 6] Corollary 3.2 and Theorem 3.3 are proved under Assumption 6, which replaces the prior by its restriction to the bounded and centered event B_p(X). Section 3 explicitly states that this projection 'is not utilized in the algorithm,' and the claim that the projection has 'negligible difference' is asserted without proof. The contraction and misclustering guarantees therefore do not formally cover the CAVI or SVI posteriors actually computed. To close this gap, the paper needs a quantitative statement (e.g., a lower bound on Pi(B_p) or a uniform bound showing the variational ELBO is asymptotically unaffected by the projection) rather than an appeal to a remark in Ma et al. (2020).
  3. [Section 3, Theorem 3.3] Theorem 3.3 says 'Suppose the assumptions in Theorem 3.2 hold,' but there is no Theorem 3.2 in the manuscript. The evident intended statement is Corollary 3.2 (or a separate latent-vector convergence theorem). The theorem numbering should be fixed and all assumptions used in Theorem 3.3 should be stated explicitly.
  4. [Appendix A.2, equations (A.7) and (A.13)] The definitions of tau*_x in (A.7) and tau*_b in (A.13), and the inequalities following them, are difficult to verify: the quantities delta and tau* appear to have different implied scales (a factor of n appears in some denominators and not in others), and the constants K, K', K'' in (A.8)-(A.10) are not tracked through the final exponent. Because the claimed prior-mass lower bound e^{-K n^2 epsilon_n^2} is the basis of Theorem 3.1, these estimates need to be rechecked carefully.
minor comments (4)
  1. [Corollary 3.2 display] The inequality in Corollary 3.2 is written as '>= MD + 3eta/(1-alpha) e^{-beta*} epsilon_n^2', which should be '>= M(D+3eta)/(1-alpha) e^{-beta*} epsilon_n^2'; the missing parentheses make the statement ambiguous.
  2. [Section 2.1, equation (3)] The prior for x_i should specify the covariance as lambda_{xi}^2 tau_x^2 I_d rather than just lambda_{xi}^2 tau_x^2 for dimensional consistency.
  3. [Figure 2 caption and Section 5.1] The caption of Figure 2 says 'SVD performs the second best' in weak-signal settings, while the text says SVDyzO is second best; this inconsistency should be resolved.
  4. [Section 3 and Lemma A.5] There are typesetting and terminology issues: 'Asumption 5' and 'indetifibility' should be corrected, and Lemma A.5 uses h^2 without defining it; the standard Hellinger affinity should be defined.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: contraction rates are derived from prior-mass computations and external general theory; self-citations are not load-bearing.

full rationale

The derivation chain is a standard prior-to-posterior contraction argument. Theorem 3.1 is proved by lower-bounding the prior mass of a KL/second-moment neighborhood of the truth (Appendix A.2) and then invoking Theorem 3.2 of Bhattacharya et al. (2019); the target rate epsilon_n^2 is the radius at which the horseshoe prior is shown to concentrate, not a quantity fitted to data or imposed as the conclusion. Assumption 4 is a prior-mass condition on beta, and Lemma A.2 verifies it for a N(0,log n) prior; it does not assume the contraction result. Corollary 3.2 and Theorem 3.3 follow from Theorem 3.1 via Lemmas A.3-A.7, which are either elementary or quoted from external sources (Tu et al. 2016; Lei and Rinaldo 2015); no fitted parameter is renamed as a prediction. Self-citations (Zhao et al. 2022a,b; Niu et al. 2023) appear only in the introduction, related-work discussion, and algorithmic motivation, not as proof of the contraction or community-detection results. Assumption 6's projection is explicitly stated to be 'used solely to simplify proving the theorem and is not utilized in the algorithm'; this is an honest coverage gap between the theory and the implemented variational algorithms, not a circular reduction. The possible sqrt(n) scaling concern in Theorem 3.3's proof is a mathematical correctness issue: Corollary 3.2 gives ||Xhat Xhat' - X* X*'||_F of order n e^{-beta*/2} epsilon_n, and Appendix A.4 divides by sqrt(n) to obtain ||Xhat - X*O||_F of order e^{-beta*/2} epsilon_n, dropping the extra factor n. This does not reduce the claimed result to its inputs by definition, so it is not circularity. Overall, the paper's central claims are self-contained given its stated assumptions, and I find no significant circularity.

Assumptions & free parameters 5 free parameters · 9 assumptions · 0 invented entities

The theory rests on six modeling assumptions, plus external fractional posterior and matrix perturbation results. No new physical or probabilistic entities are invented; latent positions and regression coefficients are standard model parameters. The main load-bearing free choices are the latent dimension, the SVI initialization hyperparameters, and the unspecified fractional posterior power used in the algorithms.

free parameters (5)
  • Latent dimension d = d = 5 for Cora, Yeast, LastFM100; d = 3 for PubMed
    Chosen by the user; no selection criterion is given in the paper.
  • SVI Gamma initialization hyperparameters = alpha_lambda_x = alpha_lambda_b = 10.0; beta_lambda_x = beta_lambda_b = 10.0; alpha_tau_x = alpha_tau_b = 0.1…
    Selected through 'systematic experimentation' in Section 4.2; these affect the sparsity behavior of the variational posterior.
  • Negative edge sampling ratio = 5 negative edges per positive edge
    Described in Section 4.2 as empirically optimal for the accuracy-efficiency trade-off.
  • Optimizer and scheduling parameters = AdamW, learning rate 0.005, weight decay 1e-4, batch sizes 128 to 2048, S = 10 Monte Carlo samples
    Set in Section 5.2 for real-data SVI experiments; the paper says other parameters were kept consistent across datasets.
  • Fractional posterior power alpha = unspecified in the algorithm sections
    Theory requires alpha in (0,1), but the CAVI and SVI implementations do not state which alpha is used in the experiments.
assumptions (9)
  • domain assumption Assumption 1: true latent positions are bounded in l_infinity and covariate coefficients grow at most polynomially.
    Used throughout the proof to control inner products and prior concentration; stated in Section 3.
  • domain assumption Assumption 2: restricted eigenvalue condition on Z over the sparse cone.
    Needed to transfer l_{2,1} error in B to Frobenius error in Z(B-B*); stated in Section 3.
  • domain assumption Assumption 3: edge probability is not too small, e^{beta*} much larger than epsilon_n^2.
    Required for consistent recovery of inner products and for the community detection rate; stated in Section 3.
  • domain assumption Assumption 4: the prior on beta has enough mass around beta*.
    A standard prior mass condition for fractional posterior contraction; satisfied by N(0, log n) as shown in Lemma A.2.
  • domain assumption Assumption 5: true latent positions sum to zero for identifiability between X and beta.
    Used in Corollary 3.2 to control cross terms; stated in Section 3.
  • ad hoc to paper Assumption 6: the prior is restricted to bounded and centered latent vectors.
    Stated as a proof simplification that is not utilized in the algorithm; this creates a gap between theory and implementation.
  • domain assumption The data generating process is a latent space model with observed diagonals and known latent dimension d.
    The likelihood and theoretical analysis assume Y_ij ~ Bernoulli(logistic(beta* + x_i' x_j)), including diagonal entries of Y.
  • standard math External fractional posterior contraction theorem of Bhattacharya et al. (2019).
    Invoked in the proof of Theorem 3.1 to turn prior mass conditions into posterior contraction.
  • standard math Procrustes perturbation bound (Tu et al. 2016) and K-means error bound (Lei and Rinaldo 2015).
    Used in the proofs of Theorem 3.3 to convert latent vector estimation error into misclustering error.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Robust High-Dimensional Covariate-Assisted Network Modeling." pith.science (2026). https://pith.science/paper/C4PFQOBE

@misc{pith2026250502986,
  author       = {Pith},
  title        = {Pith review of: Robust High-Dimensional Covariate-Assisted Network Modeling},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/C4PFQOBE}},
  note         = {Machine review of arXiv:2505.02986}
}
read the original abstract

Modern network data analysis often involves analyzing network structures alongside covariate features to gain deeper insights into underlying patterns. However, traditional covariate-assisted statistical network models may not adequately handle cases involving high-dimensional covariates, where some covariates could be uninformative or misleading, or the possible mismatch between network and covariate information. To address this issue, we introduce a novel robust high-dimensional covariate-assisted latent space model. This framework links latent vectors representing network structures with simultaneously sparse and low-rank transformations of the high-dimensional covariates, capturing the mutual dependence between network structures and covariates. To robustly integrate this dependence, we use a shrinkage prior on the discrepancy between latent network vectors and low-rank covariate approximation vectors, allowing for potential mismatches between network and covariate information. For scalable inference, we develop two variational inference algorithms, enabling efficient analysis of large-scale sparse networks. We establish the posterior concentration rate within a suitable parameter space and demonstrate how the proposed model facilitates adaptive information aggregation between networks and high-dimensional covariates. Extensive simulation studies and real-world data analyses confirm the effectiveness of our approach.

Figures

Figures reproduced from arXiv: 2505.02986 by the authors.

Figure 1
Figure 1. A toy example showing the mismatch between binary covariate values and cluster [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Comparison of community detection methods in terms of PCC [PITH_FULL_IMAGE:figures/full_fig_p024_2.png] view at source ↗
Figure 3
Figure 3. Comparison of median PCC across 50 Monte Carlo simulations for different [PITH_FULL_IMAGE:figures/full_fig_p025_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Comparison of community detection methods in terms of RI [PITH_FULL_IMAGE:figures/full_fig_p027_4.png]
Figure 5
Figure 5. Figure 5: Comparison of community detection methods in terms of RI [PITH_FULL_IMAGE:figures/full_fig_p027_5.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Testing for correlation between network structure and high-dimensional node covariates

    stat.ML 2025-09 conditional novelty 7.0 of 10

    Four permutation tests (ridge, group LASSO, and two CCA variants) detect association between node covariates and random-dot-product-graph latent structure, with consistency theorems and cheaper computation than prior ...

  2. AI-Augmented Statistical Network Estimation with Proxy Gene Embeddings

    stat.ME 2026-07 conditional novelty 6.0 of 10

    PLANE jointly estimates latent gene positions from a target network and proxy embeddings on a larger gene set, with provably optimal channel weighting and demonstrated gains in network recovery and imputation.

Reference graph

Works this paper leans on

54 extracted references · 39 canonical work pages · cited by 2 Pith papers

  1. [1]

    R., and Oliveira Jr, O

    Akimushkin, C., Amancio, D. R., and Oliveira Jr, O. N. (2017). Text authorship identified using the dynamics of word co-occurrence networks. PloS one , 12(1):e0170527

    work page 2017

  2. [2]

    Bhattacharya, A., Pati, D., and Yang, Y. (2019). Bayesian fractional posteriors. The Annals of Statistics , 47(1):39--66

    work page 2019

  3. [3]

    J., Ritov, Y., and Tsybakov, A

    Bickel, P. J., Ritov, Y., and Tsybakov, A. B. (2009). Simultaneous analysis of lasso and dantzig selector. The Annals of Statistics , pages 1705--1732

    work page 2009

  4. [4]

    T., and Rohe, K

    Binkiewicz, N., Vogelstein, J. T., and Rohe, K. (2017). Covariate-assisted spectral clustering. Biometrika , 104(2):361--377

    work page 2017

  5. [5]

    Bishop, C. M. and Nasrabadi, N. M. (2006). Pattern recognition and machine learning , volume 4. Springer

    2006

  6. [6]

    M., Kucukelbir, A., and McAuliffe, J

    Blei, D. M., Kucukelbir, A., and McAuliffe, J. D. (2017). Variational inference: A review for statisticians. Journal of the American Statistical Association , 112(518):859--877

    2017

  7. [7]

    M., Polson, N

    Carvalho, C. M., Polson, N. G., and Scott, J. G. (2010). The horseshoe estimator for sparse signals. Biometrika , 97(2):465--480

    2010

  8. [8]

    Csardi, G. (2015). igraphdata: A Collection of Network Data Sets for the 'igraph' Package . R package version 1.0.1

    work page 2015

Show all 54 references
  1. [9]

    and Fran c ois, R

    Eddelbuettel, D. and Fran c ois, R. (2011). Rcpp : Seamless R and C++ integration. Journal of Statistical Software , 40(8):1--18

    2011

  2. [10]

    and Hric, D

    Fortunato, S. and Hric, D. (2016). Community detection in networks: A user guide. Physics reports , 659:1--44

    2016

  3. [11]

    X., Fienberg, S

    Goldenberg, A., Zheng, A. X., Fienberg, S. E., and Airoldi, E. M. (2010). A survey of statistical network models

    2010

  4. [12]

    Hager, G. L. and Talbert, J. C. (2000). Look for the party label: Party influences on voting in the us house. Legislative Studies Quarterly , pages 75--99

    2000

  5. [13]

    S., Raftery, A

    Handcock, M. S., Raftery, A. E., and Tantrum, J. M. (2007). Model-based clustering for social networks. Journal of the Royal Statistical Society: Series A (Statistics in Society) , 170(2):301--354

    2007

  6. [14]

    Hoff, P. (2008). Modeling homophily and stochastic equivalence in symmetric relational data. In Platt, J., Koller, D., Singer, Y., and Roweis, S., editors, Advances in Neural Information Processing Systems , volume 20, pages 657--664. Curran Associates, Inc

    2008

  7. [15]

    D., Raftery, A

    Hoff, P. D., Raftery, A. E., and Handcock, M. S. (2002). Latent space approaches to social network analysis. Journal of the American Statistical Association , 97(460):1090--1098

    2002

  8. [16]

    and Wang, W

    Hu, Y. and Wang, W. (2024). Network-adjusted covariates for community detection. Biometrika , page asae011

    2024

  9. [17]

    Jaakkola, T. S. and Jordan, M. I. (2000). Bayesian parameter estimation via variational methods. Statistics and Computing , 10(1):25--37

    2000

  10. [18]

    and Ghosal, S

    Jeong, S. and Ghosal, S. (2021). Posterior contraction in sparse generalized linear models. Biometrika , 108(2):367--379

    2021

  11. [19]

    M., and Jordan, M

    Jin, C., Netrapalli, P., Ge, R., Kakade, S. M., and Jordan, M. I. (2019). A short note on concentration inequalities for random vectors with subgaussian norm. arXiv preprint arXiv:1902.03736

    2019 arXiv

  12. [20]

    N., Handcock, M

    Krivitsky, P. N., Handcock, M. S., Raftery, A. E., and Hoff, P. D. (2009). Representing degree distributions, clustering, and homophily in social networks with latent cluster random effects models. Social networks , 31(3):204--213

    2009

  13. [21]

    and Rinaldo, A

    Lei, J. and Rinaldo, A. (2015). Consistency of spectral clustering in stochastic block models. The Annals of Statistics , 43(1):215--237

    2015

  14. [22]

    Le (2023)

    Li, T., Levina, E., Zhu, J., and Can M. Le (2023). randnet: Random Network Model Estimation, Selection and Parameter Tuning . R package version 0.7

    2023

  15. [23]

    and Chen, Y

    Liu, Y. and Chen, Y. (2022). Variational inference for latent space models for dynamic networks. Statistica Sinica , 32(4):2147--2170

    2022

  16. [24]

    and Hutter, F

    Loshchilov, I. and Hutter, F. (2019). Decoupled weight decay regularization. In International Conference on Learning Representations

    2019

  17. [25]

    Loyal, J. D. (2024). Fast variational inference of latent space models for dynamic networks using Bayesian p-splines. arXiv preprint arXiv:2401.09715

    2024 arXiv

  18. [26]

    Loyal, J. D. and Chen, Y. (2023). An eigenmodel for dynamic multilayer networks. The Journal of Machine Learning Research , 24(128):1--69

    2023

  19. [27]

    Ma, Z., Ma, Z., and Yuan, H. (2020). Universal latent space model fitting for large networks with edge covariates. The Journal of Machine Learning Research , 21(4):1--67

    2020

  20. [28]

    and Tang, Y

    Martin, R. and Tang, Y. (2020). Empirical priors for prediction in sparse high-dimensional linear regression. The Journal of Machine Learning Research , 21(1):5709--5738

    2020

  21. [29]

    Mart \' nez, V., Berzal, F., and Cubero, J.-C. (2016). A survey of link prediction in complex networks. ACM computing surveys (CSUR) , 49(4):1--33

    2016

  22. [30]

    K., Nigam, K., Rennie, J., and Seymore, K

    McCallum, A. K., Nigam, K., Rennie, J., and Seymore, K. (2000). Automating the construction of internet portals with machine learning. Information Retrieval , 3:127--163

    2000

  23. [31]

    Mitchell, T. J. and Beauchamp, J. J. (1988). Bayesian variable selection in linear regression. Journal of the american statistical association , 83(404):1023--1032

    1988

  24. [32]

    Namata, G., London, B., Getoor, L., Huang, B., and Edu, U. (2012). Query-driven active surveying for collective classification. In 10th international workshop on mining and learning with graphs , volume 8, page 1

    2012

  25. [33]

    E., Ormerod, J

    Neville, S. E., Ormerod, J. T., and Wand, M. (2014). Mean field variational Bayes for continuous sparse signal shrinkage: pitfalls and remedies. Electronic Journal of Statistics , 8(1):1113--1151

    2014

  26. [34]

    Newman, M. (2018). Networks . Oxford university press

    2018

  27. [35]

    Newman, M. E. and Clauset, A. (2016). Structure and inference in annotated networks. Nature communications , 7(1):11863

    2016

  28. [36]

    Niu, Y., Ni, Y., Pati, D., and Mallick, B. K. (2023). Covariate-assisted Bayesian graph learning for heterogeneous data. Journal of the American Statistical Association , pages 1--15

    2023

  29. [37]

    and George, E

    Rockova, V. and George, E. I. (2018). The spike - and - slab lasso. Journal of the American Statistical Association , 113(521):431--444

    2018

  30. [38]

    Rowe, R., Creamer, G., Hershkop, S., and Stolfo, S. J. (2007). Automated social hierarchy detection through email network analysis. In Proceedings of the 9th WebKDD and 1st SNA-KDD 2007 workshop on Web mining and social network analysis , pages 109--117

    2007

  31. [39]

    and Sarkar, R

    Rozemberczki, B. and Sarkar, R. (2020). Characteristic Functions on Graphs: Birds of a Feather, from Statistical Descriptors to Parametric Models . In Proceedings of the 29th ACM International Conference on Information and Knowledge Management (CIKM '20) , page 1325–1334. ACM

    2020

  32. [40]

    Sewell, D. K. and Chen, Y. (2017). Latent space approaches to community detection in dynamic networks. Bayesian analysis , 12(2):351--377

    2017

  33. [41]

    Snijders, T. A. (2011). Statistical models for social networks. Annual review of sociology , 37:131--153

    2011

  34. [42]

    and Lehmann, S

    Stopczynski, A. and Lehmann, S. (2018). How physical proximity shapes complex social networks. Scientific reports , 8(1):1--10

    2018

  35. [43]

    Tu, S., Boczar, R., Simchowitz, M., Soltanolkotabi, M., and Recht, B. (2016). Low-rank solutions of linear matrix equations via procrustes flow. In International Conference on Machine Learning , pages 964--973. PMLR

    2016

  36. [44]

    Vershynin, R. (2010). Introduction to the non-asymptotic analysis of random matrices. arXiv preprint arXiv:1011.3027

    2010 arXiv

  37. [45]

    G., Fields, S., and Bork, P

    Von Mering, C., Krause, R., Snel, B., Cornell, M., Oliver, S. G., Fields, S., and Bork, P. (2002). Comparative assessment of large-scale data sets of protein--protein interactions. Nature , 417(6887):399--403

    2002

  38. [46]

    and Hjort, N

    Walker, S. and Hjort, N. L. (2001). On Bayesian consistency. Journal of the Royal Statistical Society: Series B (Statistical Methodology) , 63(4):811--821

    2001

  39. [47]

    and Sarkar, P

    Yan, B. and Sarkar, P. (2021). Covariate regularized community detection in sparse graphs. Journal of the American Statistical Association , 116(534):734--745

    2021

  40. [48]

    E., and Leng, C

    Yan, T., Jiang, B., Fienberg, S. E., and Leng, C. (2019). Statistical inference in a directed network model with covariates. Journal of the American Statistical Association , 114(526):857--868

    2019

  41. [49]

    Zhang, J., He, X., and Wang, J. (2022a). Directed community detection with network embedding. Journal of the American Statistical Association , 117(540):1809--1819

    2022

  42. [50]

    Zhang, X., Xu, G., and Zhu, J. (2022b). Joint latent space models for network data with high-dimensional node variables. Biometrika , 109(3):707--720

    2022

  43. [51]

    Zhang, X., Xue, S., and Zhu, J. (2020). A flexible latent space model for multilayer networks. In International Conference on Machine Learning , pages 11288--11297. PMLR

    2020

  44. [52]

    Zhang, Y., Levina, E., and Zhu, J. (2016). Community detection in networks with node features. Electronic Journal of Statistics , 10(2):3153--3178

    2016

  45. [53]

    Zhao, P., Bhattacharya, A., Pati, D., and Mallick, B. K. (2022a). Factorized fusion shrinkage for dynamic relational data. arXiv preprint arXiv:2210.00091

    2022 arXiv

  46. [54]

    Zhao, P., Bhattacharya, A., Pati, D., and Mallick, B. K. (2022b). Structured optimal variational inference for dynamic latent space models. arXiv preprint arXiv:2209.15117

    2022 arXiv

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.