Graph Topic Modeling for Documents with Spatial or Covariate Dependencies

Claire Donnat; Yeo Jin Jung

REVIEW 4 major objections 6 minor 8 references

Graph Topic Modeling for Documents with Spatial or Covariate Dependencies

T0 review · 4 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read Graph-regularized pLSI provably improves topic estimates when documents are linked by a known graph.

desk verdict Useful graph-regularized pLSI estimator with strong empirical work, but the main W-rate theorem is not actually established because the proof's small-ρ condition contradicts the oracle penalty choice. read the letter →

arxiv 2412.14477 v2 pith:DT5QA4HI submitted 2024-12-19 cs.LG stat.ME

Yeo Jin Jung , Claire Donnat This is my paper · ORCID

classification cs.LGstat.ME MSC 62H3062F12

keywords topicmodelingprobabilisticlatentsemanticindexinggraphregularizationtotalvariationpenaltysingularvaluedecompositionspatialtranscriptomicscross-validationmixtureestimation

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 GpLSI, a frequentist topic model that uses a known similarity graph between documents to sharpen estimates of the document-topic mixture matrix $W$. The core claim is that replacing the plain SVD in pLSI with an iterative, graph-regularized SVD—one that penalizes differences in topic proportions across neighboring documents—reduces estimation error provably, with the gain largest for short documents. Under the paper's assumptions, the Frobenius error of $W$ is bounded by $C K \sqrt{\log(n)/N}\left(\sqrt{n_C} + \rho(\Gamma)\sqrt{s\lambda_{\max}(\Gamma)}\right)$, improving on the pLSI rate whenever the graph is informative. Because the method is frequentist, it avoids the computational cost of Bayesian spatial LDA while providing high-probability guarantees. The authors demonstrate the practical benefit on synthetic data and three real corpora: two spatial transcriptomics datasets and a cuisine/recipe dataset.

What carries the argument

The load-bearing object is the graph incidence matrix $\Gamma$ (with Laplacian $L = \Gamma^{\top}\Gamma$) and the total-variation penalty $\|\Gamma U\|_{2,1}$ used in each iteration of Algorithm 1. The iterative scheme alternates: denoise the left singular vectors of $X\hat{V}^{t-1}$ under this penalty, take a rank-$K$ SVD, then update $\hat{V}$ from $X^{\top}\hat{U}^t$; the penalty's effect is controlled by the inverse scaling factor $\rho(\Gamma)$, the number of connected components $n_C$, and the support size $s$ of $\Gamma W$. The same penalty, together with a minimum-spanning-tree cross-validation (Algorithm 2) for the regularization parameter, is what lets neighboring documents share statistical strength. The proof chain adapts the unregularized pLSI analysis to these denoised singular vectors via robustness of successive projections and singular-subspace perturbation bounds.

What would settle it

Take a grid graph with $n = 10^4$ nodes, $K = 3$ topics, document length $N = 30$, vocabulary size $p = 50$, and generate $W$ exactly piecewise-constant with known support size $s$ for $\Gamma W$. Run Algorithm 1 with the oracle penalty $\rho = 4C^*\rho(\Gamma)\sqrt{K\log(n)/N}(1+L_{t-1})$ and measure $\min_{P \in \mathcal{P}} \|\hat{W} - W P\|_F$. If the error does not decrease when $s$ shrinks, or if it exceeds the right-hand side of the bound in Theorem 3 for a large enough constant $C$, the theorem would be contradicted.

Watch

Extended reading notes

Core claim

The central discovery is that document-level metadata, encoded as edges of a graph, can be folded into pLSI without giving up either speed or theory. GpLSI iteratively computes singular vectors of the word-frequency matrix while solving a total-variation-denoised update on the left singular subspace, then recovers $W$ by vertex hunting on the denoised simplex. Theorem 3 states that with high probability $\min_{P \in \mathcal{P}} \|\hat{W} - W P\|_F \le C K \sqrt{\log(n)/N}\left(\sqrt{n_C} + \rho(\Gamma)\sqrt{s\lambda_{\max}(\Gamma)}\right)$, under a document-length condition that is milder than the $N \gtrsim \log(n+p)$ needed by unregularized pLSI. The factor $\left(\sqrt{n_C} + \rho(\Gamma)\sqrt{s\lambda_{\max}(\Gamma)}\right)/\sqrt{n}$ is the paper's measure of how much the graph helps: it equals one for an empty graph and shrinks for connected structured graphs such as grids.

Load-bearing premise

The high-probability theorems assume the regularization penalty is set to an oracle value that depends on unknown noise and graph quantities, while the implemented algorithm chooses it by cross-validation; nothing in the paper proves the two choices are close enough for the bounds to apply to the actual method.

Editorial extensions

If this is right

For well-connected bounded-degree graphs such as grids, k-nearest-neighbor graphs, and spatial graphs, the $W$-error bound improves by roughly a factor of $(1 + \sqrt{s\log n})/\sqrt{n}$ over unregularized pLSI, so short documents become usable in practice.
When the graph is empty or carries no signal, the bound reduces to the pLSI rate, so the method does not hurt in the absence of metadata.
A corollary gives an $\ell^1$ error bound for $\hat{W}$, and the error bound for the topic matrix $\hat{A}$ is governed by the accuracy of $\hat{W}$, making mixture recovery the primary target.
The required document length $N$ is relaxed relative to unregularized pLSI, which is the regime where the graph's information matters most.
The minimum-spanning-tree cross-validation gives a data-driven choice of the graph penalty, and the paper observes that the chosen penalty shrinks as $N$ grows, consistent with the theory's diminishing need for smoothing.

Reading between the lines

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

Extension (ours): the same iterative graph-aligned SVD could serve as a generic denoiser for low-rank matrix estimation under row-smoothness constraints, beyond topic models—for example, collaborative filtering with user-item graphs—since the proof uses only the signal-plus-noise structure of the data matrix.
Extension (ours): the paper's Remark 2 notes that a rare-word thresholding step could lift the low-vocabulary restriction; if that extension holds, the graph gain would apply to genomics-scale vocabularies, a setting the paper does not test.
Extension (ours): a direct empirical comparison between the cross-validation-selected penalty and the oracle penalty used in the proofs would clarify whether the implemented algorithm attains the theoretical rates; this is a testable check, not a claim of failure.

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.

Desk Editor's Note

read the letter

Colleague, here's my read on Jung and Donnat's GpLSI paper. The core idea is solid and timely: take pLSI's SVD-plus-SPA pipeline and add a graph total-variation penalty on the left singular vectors so that neighboring documents share similar topic mixtures. That is a genuinely new estimator, and the paper backs it with an iterative algorithm, a specialized MST-based cross-validation, and a serious experimental section on CODEX and recipe data. Code is public. The authors also explicitly flag the low-p limitation and point to a sparsity-based extension, which is honest framing. Credit where due: the paper does what a good methods paper should do—it identifies a real problem, proposes a plausible fix, and tests it across multiple settings.

The soft spots are real, and they are in the theory. The reader's concern about the oracle penalty is correct but understated. Lemma 1 sets ρ to an oracle value, while Algorithm 2 tunes ρ via MST-CV; there is no result showing the CV choice is close enough to the oracle, so the contraction argument in Theorem 2 does not automatically apply to the implemented algorithm. That is a gap, but it could be fixable.

The stress-test note is more serious, and I think it lands. In the proof of Theorem 3 (Appendix D), the row-wise bound for ŽU requires ρ to be a small value satisfying an upper bound that is essentially O(√(K log n/N)·(√n_C+ρ(Γ)√(sλ_max))/d_max). The oracle ρ from Lemma 1 is of order ρ(Γ)√(K log n/N), which is larger by roughly a factor of √n in the grid-graph setting. So the SPA-based W-rate in Theorem 3 is not proven for the estimator whose singular-vector contraction was established in Theorem 2. This is not an error in a minor lemma; it cuts the load-bearing theoretical claim. The empirical results may well still hold—and I suspect they do, given the systematic experiments—but the central theorem as stated is currently unsupported.

One more, smaller point: the real-data recipe application has p=1019, which is far outside the proved low-p regime. The authors acknowledge this in Remark 2, but the leap from theory to application is bigger than the paper's tone suggests.

Verdict: this deserves a serious referee, but with the clear expectation of major revision. The method is worth publishing after the theory is either fixed (perhaps by a different penalty scaling or a two-stage argument) or honestly narrowed to a Frobenius-norm bound that does not rely on the small-ρ condition. The empirical contribution alone is strong enough to justify a revise-and-resubmit.

I would not cite the current version's rates in my own work, but I would cite the method and the experiments once they are cleaned up.

Referee Report

4 major / 6 minor

Summary. This paper proposes GpLSI, a frequentist topic model that incorporates document-level metadata encoded as a graph by replacing the rank-K SVD step of pLSI with an iterative graph-regularized SVD. The authors derive high-probability error bounds for the estimated singular vectors, for the topic-mixture matrix W (Theorem 3), and for the topic-word matrix A (Theorem 4), with rates involving the graph's inverse scaling factor, the number of connected components, and the support size of the edge differences. They also introduce an MST-based cross-validation procedure for the regularization parameter and validate the method on synthetic data and three real-world datasets.

Significance. If the central rate in Theorem 3 is correct, this is the first theoretical guarantee for graph-structured pLSI and it shows a clear improvement over the graph-free bound of Klopp et al. in piecewise-smooth, short-document regimes. The paper's strengths include the extensive proof machinery, the reproducible code, the systematic synthetic evaluation showing gains in the short-document regime, and the real-data demonstrations. The significance is currently tempered by a proof-level gap in Theorem 3 and by the mismatch between the oracle penalty assumed in the theory and the CV-based penalty used in the implementation.

major comments (4)

[Appendix D, after Eq. (44)] The proof of Theorem 3 imposes an unstated small-ρ condition: the row-wise bound on U is claimed only 'when ρ is a small value that satisfies ρ ≤ (1/λ_K(M)) sqrt(K log n/N) (√n_C + ρ(Γ)√(sλ_max))/d_max.' This condition is not stated in Theorem 3, and it is violated by the oracle penalty of Lemma 1, ρ = 4C*ρ(Γ)√(K log n/N)(1+L_{t-1}). For a 2D grid (ρ(Γ) ≲ √log n, d_max ≤ 4, n_C = 1), the oracle penalty is of order √(K log^2 n/N), while the required upper bound is of order (K/√n)√(log n/N)(1+√(s log n))/4, which is smaller by a factor of about √n/√log n. Thus the SPA row-wise analysis does not apply to the estimator whose singular-vector contraction was proved in Theorem 2, and the W-estimation rate in Theorem 3 is not established by the supplied argument.
[Section 2.3 and Theorems 2–4; Lemma 1, Appendix C.1] The theoretical results assume an oracle choice of the graph penalty, ρ = 4C*ρ(Γ)√(K log n/N)(1+L_{t-1}), while Algorithm 1 as implemented selects ρ_t via the MST cross-validation of Algorithm 2. No result shows that the CV-selected ρ_t is close to the oracle value or that the error contraction of Theorem 2 survives data-dependent tuning. As stated, the high-probability bounds apply to a variant of GpLSI that is not the one evaluated in the experiments; the paper should either provide a theoretical analysis of the CV selection or explicitly separate the oracle-based theory from the CV-based implementation.
[Appendix C.1, proof of Theorem 2] The bound on ∥sin Θ(V, bV^t)∥F uses the estimate ∥Z^T U⊥∥_op ≤ C√(Kn log n/N), citing Lemma 15. However, Lemma 15 is stated for a matrix U ∈ R^{n×r} with r not growing with n, whereas U⊥ has r = n − K, which grows with n. The stated lemma therefore does not cover the operator norm bound used here, and no other lemma in the paper justifies it. This is a load-bearing step in the contraction argument for the right singular vectors; please supply a proof or a precise reference for this estimate.
[Section 4.3 and Remark 2] The 'What's Cooking' experiment uses a vocabulary of p = 1,019 ingredients with short documents, which is far outside the low-p regime enforced by Assumption 5: as Remark 2 notes, that assumption implies p ≤ KN/(c_min log n). The thresholding remedy suggested in Remark 2 is not applied in this experiment. The paper should either apply that thresholding or explicitly state that the theoretical guarantees do not cover this experiment, and should temper the claim that the experiments validate the theory on this dataset.

minor comments (6)

[Section 2.2] The sentence 'Mij denotes the expected frequency of word i in document j' should read 'word j in document i', since M is an n × p document-word matrix.
[Appendix B, proof of Theorem 1] There is a stray 'E' in the line 'We have: Z^T Z = ... E', which appears to be a typographical artifact and should be removed.
[Figure 2 caption] The caption contains a typo: 'Errors are normalzied by n' should be 'Errors are normalized by n'.
[Algorithm 1] The pseudocode uses ρ̂_t in the graph-denoising step, but the input list of Algorithm 1 does not include ρ̂_t or the cross-validation procedure that selects it; the pseudocode should be harmonized with the surrounding text.
[Section 3.3, Eq. (17)] The displayed bound for Erdős-Rényi graphs is garbled by a line break: '1 + s 1 2 log(n)−3 4' is not readable as a mathematical expression. Please restate the formula cleanly.
[Section 2.1] The assumption 'n_{C1} ≍ ... ≍ n_{Cn_C}' is stated but does not appear to be used in the proofs; Lemma 14, for instance, does not require equal connected-component sizes. Consider removing the assumption or explaining where it is needed.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the GpLSI error bounds are derived from stated model assumptions and concentration arguments, and the only self-citation is used for a proposed extension and an evaluation metric, not for the central derivation.

full rationale

The paper's central claims, Theorems 2-4, are derived from explicit assumptions (graph-piecewise W, anchor documents, equal document sizes, spectral conditions, low-p word frequency), from concentration inequalities for multinomial noise, and from standard singular-subspace and SPA perturbation results. The W-estimation bound is obtained by feeding the graph-denoised singular-vector error into the SPA vertex-hunting analysis, and the A-bound follows by regressing X on the estimated W; neither bound is fitted to the target quantities through the paper's own equations. The cross-validation procedure in Algorithm 2 selects the regularization parameter from data, whereas the theorems use an oracle penalty value; this is a mismatch between implemented and analyzed procedures, and is a proof-level concern, not a fitted-input-called-prediction circularity. Likewise, the Appendix D small-rho condition required for the two-to-infinity row-wise bound is not checked against the oracle rho of Lemma 1; this is an unverified proof condition, not a circular reduction. The only self-citation is Tran, Liu & Donnat (2023), which appears in Remark 2 as a proposed thresholding extension to lift the low-p limitation and as a suggested batch-coherence evaluation metric in Section 4.1; neither use is load-bearing for the main theorems or for the experimental comparisons, which are evaluated against external baselines (pLSI, LDA, TopicSCORE, SLDA). No step was found where a prediction reduces by construction to a fitted parameter, a self-citation chain, or a definitional equivalence.

Assumptions & free parameters 4 free parameters · 8 assumptions · 0 invented entities

GpLSI relies on the standard pLSI data model plus five explicit assumptions (graph-aligned mixtures, anchor documents, equal lengths, condition numbers, low-p word frequency). The main tuned inputs are graph construction, K, and the CV-selected TV penalty; no new particles, mediators, or latent entities are introduced. The low-p assumption and the oracle-versus-CV penalty gap are the costliest entries.

free parameters (4)

graph construction (neighborhood size, edge weights) = 5 nearest neighbors with inverse-distance weights in synthetic and recipe data; 3-hop Voronoi neighborhood in CODEX…
The document graph is user-specified and strongly influences the regularization path; no data-driven rule for graph construction is given.
graph regularization parameter rho_t = selected per iteration by MST cross-validation in practice; oracle value in theory is 4 C* rho(Gamma) sqrt(K log n /…
Controls the total-variation penalty; the theoretical choice depends on unknown constants and graph quantities.
number of topics K = set per dataset, e.g., K = 6 for CODEX and K = 7 for recipes; chosen via topic path and coherence heuristics
K is a model selection parameter fixed before fitting; it is not estimated by the algorithm.
cross-validation fold count b and SVD convergence tolerance epsilon = not reported
Algorithm 2 takes b folds and Algorithm 1 takes tolerance epsilon; defaults are not stated, which affects reproducibility.

assumptions (8)

domain assumption Multinomial word sampling with M = W A and centered noise Z = X - M
Equations (1)-(2); the entire signal-plus-noise analysis is built on this data model.
domain assumption Assumption 1: support of Gamma W is small (piecewise-constant topic mixtures over the graph)
Section 2.1, Equation (4); the TV penalty and error bound depend on s.
domain assumption Assumption 2: each topic has an anchor document (an extreme row of W)
Section 2.1, Assumption 2; required for identifiability and for SPA vertex hunting.
domain assumption Assumption 3: equal or comparable document lengths
Section 2.1, Assumption 3; simplifies the multinomial noise covariance; the relaxation to comparable lengths is stated but not analyzed in detail.
domain assumption Assumption 4: condition number bounds on M and W
Section 2.1, Assumption 4; used throughout to convert spectral gaps into constants.
domain assumption Assumption 5: minimum word frequency h_j >= c_min log(n)/N
Section 2.1, Assumption 5 and Remark 2; forces the low-p regime and is used in Bernstein concentration, union bounds, and Lemmas 11-17.
domain assumption Known binary graph G with incidence matrix Gamma and Laplacian L = Gamma^T Gamma
Section 2.1; the graph is given as metadata, and spectral quantities rho(Gamma) and lambda_max(Gamma) enter all bounds.
standard math Standard results: Bernstein inequalities, Davis-Kahan/Wedin/Cai-Zhang singular subspace perturbation, Hutter-Rigollet TV rates, Gillis-Vavasis SPA robustness, and Klopp et al. pLSI bounds
Appendix E and Sections C-D invoke these without proof; they are external to the paper's contribution.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Graph Topic Modeling for Documents with Spatial or Covariate Dependencies." pith.science (2026). https://pith.science/paper/DT5QA4HI

@misc{pith2026241214477,
  author       = {Pith},
  title        = {Pith review of: Graph Topic Modeling for Documents with Spatial or Covariate Dependencies},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DT5QA4HI}},
  note         = {Machine review of arXiv:2412.14477}
}

read the original abstract

We address the challenge of incorporating document-level metadata into topic modeling to improve topic mixture estimation. To overcome the computational complexity and lack of theoretical guarantees in existing Bayesian methods, we extend probabilistic latent semantic indexing (pLSI), a frequentist framework for topic modeling, by incorporating document-level covariates or known similarities between documents through a graph formalism. Modeling documents as nodes and edges denoting similarities, we propose a new estimator based on a fast graph-regularized iterative singular value decomposition (SVD) that encourages similar documents to share similar topic mixture proportions. We characterize the estimation error of our proposed method by deriving high-probability bounds and develop a specialized cross-validation method to optimize our regularization parameters. We validate our model through comprehensive experiments on synthetic datasets and three real-world corpora, demonstrating improved performance and faster inference compared to existing Bayesian methods.

Figures

Figures reproduced from arXiv: 2412.14477 by the authors.

**Figure 1.** ℓ2 error for the estimator Wc (defined as minP ∈P 1 n ∥Wc − W P∥F ) for different combinations of document length N and vocabulary size p. Here, n = 1000 and K = 3. document length N is small compared to vocabulary size p, we check the errors for different combinations of N = 10, 30, 50, 100, 200, 1000 and p = 20, 30, 50, 100, 200, 500. Results [PITH_FULL_IMAGE:figures/full_fig_p017_1.png] view at source ↗

**Figure 2.** ℓ2 error of W (left) and A (middle) and computation time (right) for different corpus size n and number of topics K. Here, N = 30 and p = 30. Errors are normalzied by n. 4 Real-World Experiments To highlight the applicability of our method, we deploy it on three real-life examples. All the code for the experiments is openly accessible at https://github.com/yeojin-jung/ GpLSI. 4.1 Tumor Microenvironment discovery: th… view at source ↗

**Figure 3.** (A) Estimated tumor-immune topic weights of GpLSI, pLSI, and LDA. Topic [PITH_FULL_IMAGE:figures/full_fig_p019_3.png] view at source ↗

Figures from the paper (18 more)

**Figure 4.** Figure 4: (A) AUC for predicting cancer recurrence using isometric log-ratio transformed [PITH_FULL_IMAGE:figures/full_fig_p020_4.png]

**Figure 5.** Figure 5: (A) Visualization of estimated B cell microenvironment topics for [PITH_FULL_IMAGE:figures/full_fig_p021_5.png]

**Figure 6.** Figure 6: (A) Estimated anchor ingredients for each topic using GpLSI. (B) Proportion of [PITH_FULL_IMAGE:figures/full_fig_p023_6.png]

**Figure 7.** Figure 7: Behavior of ˆρMST −CV as N increases (left). Behavior of ℓ2 error over graph heterogeneity (right). Graph heterogeneity is characterized by ngrp, the number of patches of documents across the unit square. Each patch is assigned similar topic mixture weights [PITH_FULL…

**Figure 8.** Figure 8: Heatmap of generated ground truth W·1, W·2, W·3, representing each topic mixture weight [PITH_FULL_IMAGE:figures/full_fig_p072_8.png]

**Figure 9.** Figure 9: ℓ2 error for the estimator Ab (defined as minP ∈P 1 p ∥Ab − P A∥F ) for different combinations of document length N and vocabulary size p. Here, n = 1000 and K = 3. 3. Generate frequency matrix X We obtain the ground truth M = W A and sample each row of D from Multino…

**Figure 10.** Figure 10: ℓ1 error for the estimator Wc (defined as minP ∈P 1 n ∥Wc − W P∥11) for different combinations of document length N and vocabulary size p. Here, n = 1000 and K = 3 [PITH_FULL_IMAGE:figures/full_fig_p073_10.png]

**Figure 11.** Figure 11: ℓ1 error for the estimator Ab (defined as minP ∈P 1 p ∥Ab − P A∥11) for different combinations of document length N and vocabulary size p. Here, n = 1000 and K = 3. 73 [PITH_FULL_IMAGE:figures/full_fig_p073_11.png]

**Figure 12.** Figure 12: ℓ1 error of W (left), A (right) for different corpus size n and number of topics K. Here, N = 30 and p = 30. in particular Ke & Wang (2017), which achieves similar results to ours in [PITH_FULL_IMAGE:figures/full_fig_p074_12.png]

**Figure 13.** Figure 13: Estimated topic weights of tumor-immune microenvironments using GpLSI. [PITH_FULL_IMAGE:figures/full_fig_p075_13.png]

**Figure 14.** Figure 14: Estimated topic weights of tumor-immune microenvironments using pLSI. [PITH_FULL_IMAGE:figures/full_fig_p075_14.png]

**Figure 15.** Figure 15: Estimated topic weights of tumor-immune microenvironments using LDA. [PITH_FULL_IMAGE:figures/full_fig_p075_15.png]

**Figure 16.** Figure 16: Kaplan-Meier curves of dichotomized topic proportions using pLSI (left) and [PITH_FULL_IMAGE:figures/full_fig_p076_16.png]

**Figure 17.** Figure 17: (A) Estimated anchor ingredients for each topic using pLSI. (B) Proportion of [PITH_FULL_IMAGE:figures/full_fig_p077_17.png]

**Figure 18.** Figure 18: (A) Estimated anchor ingredients for each topic using LDA. (B) Proportion of [PITH_FULL_IMAGE:figures/full_fig_p077_18.png]

**Figure 19.** Figure 19: Top ten common words for each topic estimated by GpLSI. [PITH_FULL_IMAGE:figures/full_fig_p078_19.png]

**Figure 20.** Figure 20: Top ten common words for each topic estimated by pLSI. [PITH_FULL_IMAGE:figures/full_fig_p078_20.png]

**Figure 21.** Figure 21: Top ten common words for each topic estimated by LDA. [PITH_FULL_IMAGE:figures/full_fig_p079_21.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

8 extracted references · 6 canonical work pages

[1]

Anderson Jr, W. N. & Morley, T. D. (1985), ‘Eigenvalues of the laplacian of a graph’, Linear and multilinear algebra 18(2), 141–145. Ara´ ujo, M. C. U., Saldanha, T. C. B., Galvao, R. K. H., Yoneyama, T., Chame, H. C. & Visani, V. (2001), ‘The successive projections algorithm for variable selection in spec- troscopic multicomponent analysis’, Chemometrics...

work page 1985
[6]

J., Yalamanchili, H

Kho, S. J., Yalamanchili, H. B., Raymer, M. L. & Sheth, A. P. (2017), A novel approach for classifying gene expression data using topic modeling, in ‘Proceedings of the 8th ACM in- ternational conference on bioinformatics, computational biology, and health informatics’, pp. 388–393. Klopp, O., Panov, M., Sigalla, S. & Tsybakov, A. (2021), ‘Assigning topic...

work page 2017
[14]

& Massart, P

Boucheron, S., Lugosi, G. & Massart, P. (2013), ‘Concentration inequalities: A nonasymp- totic theory of independence oxford, uk: Oxford univ’. 79 Cai, T. T. & Zhang, A. (2018), ‘Rate-optimal perturbation bounds for singular subspaces with applications to high-dimensional statistics’. Cape, J., Tang, M. & Priebe, C. E. (2019), ‘The two-to-infinity norm an...

work page 2013
[16]

Using SVD for Topic Modeling

Feng, Y. & Lapata, M. (2010), Topic models for image annotation and text illustration, in ‘Human Language Technologies: The 2010 Annual Conference of the North American Chapter of the ACL’, Association for Computational Linguistics, pp. 831–839. Fukuyama, J., Sankaran, K. & Symul, L. (2023), ‘Multiscale analysis of count data through topic alignment’, Bio...

work page Pith review arXiv 2010
[20]

K., Young, A., Altosaar, J., Rajniak, J., Elhadad, N., Fischbach, M

Reder, G. K., Young, A., Altosaar, J., Rajniak, J., Elhadad, N., Fischbach, M. & Holmes, S. (2021), ‘Supervised topic modeling for predicting molecular substructure from mass spectrometry’, F1000Research 10, Chem–Inf. Roberts, M. E., Stewart, B. M., Tingley, D., Lucas, C., Leder-Luis, J., Gadarian, S. K., Albertson, B. & Rand, D. G. (2014), ‘Structural to...

work page 2021
[120]

(2024), Random graphs and complex networks , Vol

Van Der Hofstad, R. (2024), Random graphs and complex networks , Vol. 54, Cambridge university press. Vershynin, R. (2018), High-dimensional probability: An introduction with applications in data science, Vol. 47, Cambridge university press. Wedin, P.-˚A. (1972), ‘Perturbation bounds in connection with singular value decomposi- tion’, BIT Numerical Mathem...

arXiv 2024
[147]

Blei, D. M. & Lafferty, J. D. (2006 b), Dynamic topic models, in ‘Proceedings of the 23rd international conference on Machine learning’, pp. 113–120. Blei, D., Ng, A. & Jordan, M. (2001), ‘Latent dirichlet allocation’, Advances in neural information processing systems

work page 2001
[7203]

& Bao, H

Shao, Y., Zhou, Y., He, X., Cai, D. & Bao, H. (2009), Semi-supervised topic modeling for image annotation, in ‘Proceedings of the 17th ACM international conference on Multi- media’, pp. 521–524. Sun, D., Toh, K.-C. & Yuan, Y. (2021), ‘Convex clustering: Model, theoretical guarantee and efficient algorithm’, The Journal of Machine Learning Research 22(1), ...

arXiv 2009

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.

[1] [1]

Anderson Jr, W. N. & Morley, T. D. (1985), ‘Eigenvalues of the laplacian of a graph’, Linear and multilinear algebra 18(2), 141–145. Ara´ ujo, M. C. U., Saldanha, T. C. B., Galvao, R. K. H., Yoneyama, T., Chame, H. C. & Visani, V. (2001), ‘The successive projections algorithm for variable selection in spec- troscopic multicomponent analysis’, Chemometrics...

work page 1985

[2] [6]

J., Yalamanchili, H

Kho, S. J., Yalamanchili, H. B., Raymer, M. L. & Sheth, A. P. (2017), A novel approach for classifying gene expression data using topic modeling, in ‘Proceedings of the 8th ACM in- ternational conference on bioinformatics, computational biology, and health informatics’, pp. 388–393. Klopp, O., Panov, M., Sigalla, S. & Tsybakov, A. (2021), ‘Assigning topic...

work page 2017

[3] [14]

& Massart, P

Boucheron, S., Lugosi, G. & Massart, P. (2013), ‘Concentration inequalities: A nonasymp- totic theory of independence oxford, uk: Oxford univ’. 79 Cai, T. T. & Zhang, A. (2018), ‘Rate-optimal perturbation bounds for singular subspaces with applications to high-dimensional statistics’. Cape, J., Tang, M. & Priebe, C. E. (2019), ‘The two-to-infinity norm an...

work page 2013

[4] [16]

Using SVD for Topic Modeling

Feng, Y. & Lapata, M. (2010), Topic models for image annotation and text illustration, in ‘Human Language Technologies: The 2010 Annual Conference of the North American Chapter of the ACL’, Association for Computational Linguistics, pp. 831–839. Fukuyama, J., Sankaran, K. & Symul, L. (2023), ‘Multiscale analysis of count data through topic alignment’, Bio...

work page Pith review arXiv 2010

[5] [20]

K., Young, A., Altosaar, J., Rajniak, J., Elhadad, N., Fischbach, M

Reder, G. K., Young, A., Altosaar, J., Rajniak, J., Elhadad, N., Fischbach, M. & Holmes, S. (2021), ‘Supervised topic modeling for predicting molecular substructure from mass spectrometry’, F1000Research 10, Chem–Inf. Roberts, M. E., Stewart, B. M., Tingley, D., Lucas, C., Leder-Luis, J., Gadarian, S. K., Albertson, B. & Rand, D. G. (2014), ‘Structural to...

work page 2021

[6] [120]

(2024), Random graphs and complex networks , Vol

Van Der Hofstad, R. (2024), Random graphs and complex networks , Vol. 54, Cambridge university press. Vershynin, R. (2018), High-dimensional probability: An introduction with applications in data science, Vol. 47, Cambridge university press. Wedin, P.-˚A. (1972), ‘Perturbation bounds in connection with singular value decomposi- tion’, BIT Numerical Mathem...

arXiv 2024

[7] [147]

Blei, D. M. & Lafferty, J. D. (2006 b), Dynamic topic models, in ‘Proceedings of the 23rd international conference on Machine learning’, pp. 113–120. Blei, D., Ng, A. & Jordan, M. (2001), ‘Latent dirichlet allocation’, Advances in neural information processing systems

work page 2001

[8] [7203]

& Bao, H

Shao, Y., Zhou, Y., He, X., Cai, D. & Bao, H. (2009), Semi-supervised topic modeling for image annotation, in ‘Proceedings of the 17th ACM international conference on Multi- media’, pp. 521–524. Sun, D., Toh, K.-C. & Yuan, Y. (2021), ‘Convex clustering: Model, theoretical guarantee and efficient algorithm’, The Journal of Machine Learning Research 22(1), ...

arXiv 2009