Pith. sign in

REVIEW 1 major objections 5 minor 69 references

Generalizable Spectral Embedding with an Application to UMAP

T0 review · 1 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read The paper claims that diagonalizing a minibatch-averaged projected Laplacian recovers the rotation that separates a SpectralNet-trained eigenspace into the true graph-Laplacian eigenvectors, yielding the first spectral-embedding…

desk verdict A practical post-processing fix for SpectralNet's rotation ambiguity, with strong empirical support; the theory is thinner than the paper lets on. read the letter →

arxiv 2501.11305 v2 pith:Q4LZMAGP submitted 2025-01-20 cs.LG stat.ML

Nir Ben-Ari , Amitai Yacobi , Uri Shaham This is my paper · ORCID
classification cs.LGstat.ML MSC 68T0765F15
keywords spectralembeddingeigenvectorseparationout-of-sampleextensionscalablemethodsUMAPgraphLaplacianRayleighquotientdeeplearning
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

Spectral embedding projects data into the leading eigenspace of a graph Laplacian, but existing implementations must trade off three properties: handling new points, scaling to large datasets, and returning individual eigenvectors rather than a rotated subspace. This paper claims to break that trade-off with Sep-SpectralNet, which keeps SpectralNet's scalable and generalizable neural-network training and adds a cheap post-processing step that separates the learned eigenspace into true eigenvectors. The separation step diagonalizes a small averaged matrix to recover the rotation the network implicitly applied, and the paper shows that this separated embedding can serve as the spectral initialization for a generalizable UMAP variant called NUMAP. If correct, spectral-embedding tools such as Fiedler vectors, diffusion maps, and global-structure-preserving visualization become usable on new data points at large scale.

What carries the argument

The central object is the minibatch-averaged projected Laplacian Φ̃ = (1/T) Σ_i Y_i^T L_i Y_i, a small (k+1)×(k+1) matrix. If the trained network output satisfies Y = VQ, then each projected Laplacian equals Q^T Λ Q, so by uniqueness of eigendecomposition the eigenvectors of Φ̃ are the columns of Q^T and its eigenvalues are those of the Laplacian. The supporting identity is Lemma 1, which proves that every minimizer of the Rayleigh quotient under orthogonality is of the form VQ, so the recovered Q fully undoes the network's rotation. This diagonalization post-processing is what turns SpectralNet's eigenspace approximation into a true spectral embedding, and it is also what makes NUMAP's residual-connected UMAP network possible: only separated eigenvectors give a meaningful first-ell-dimensional spectral coordinate for the residual connection to preserve.

What would settle it

On a dataset with known ground-truth eigenvectors (for example, a two-moons or two-spheres set), compute the minibatch-averaged projected Laplacian Φ̃ for T = 1, 4, 16, and 64, then compare the recovered eigenvectors to the exact eigenvectors of the full Laplacian. If the $sin^{2}$ Grassmann distance between the recovered and true eigenvectors does not decrease toward zero as T grows, the claim that this average recovers the rotation Q is refuted.

Watch

Extended reading notes

Core claim

The paper's central claim is that the rotation-and-reflection ambiguity inherent in Rayleigh-quotient minimization—any minimizer of Tr(A^T L A) under orthogonality has the form V Q, where V is the true eigenvector matrix of the Laplacian L—can be resolved in a one-shot post-processing step. For a learned map F_θ whose output lies in the span of the first k+1 eigenfunctions, the unknown rotation Q is encoded in the projected Laplacian Φ = (VQ)^T L (VQ) = Q^T Λ Q; because Q^T is the eigenvector matrix of this small matrix, diagonalizing an estimate of it yields both the Laplacian eigenvalues and the inverse rotation needed to recover the individual eigenvectors from the network output. The paper argues that the estimate can be formed by averaging over a few random minibatches, making the separation step nearly free and preserving SpectralNet's approximately linear scaling. The result is presented as the first spectral-embedding implementation that is simultaneously generalizable, scalable, and eigenvector-separated, and as the basis for NUMAP, a generalizable UMAP that keeps the spectral initialization responsible for UMAP's global-structure preservation.

Load-bearing premise

The method assumes that averaging a small number of random minibatches approximates the full projected Laplacian well enough to recover the true rotation and eigenvectors, and the paper gives no formal guarantee for how many minibatches are needed.

Editorial extensions

If this is right

  • Sep-SpectralNet can approximate the Fiedler vector and Fiedler value for out-of-sample points in near-linear time, making algebraic connectivity analysis scalable and generalizable.
  • Diffusion Maps become generalizable: since Sep-SpectralNet approximates both the Laplacian eigenvectors and eigenvalues, the diffusion coordinates (1-λ_i)^t v_i can be computed for new points by a single feed-forward pass.
  • NUMAP, a spectral-initialized parametric UMAP, preserves global structure better than Parametric UMAP while maintaining comparable local-structure quality and enabling fast inference; the paper's ablation shows that removing eigenvector separation makes this global preservation inconsistent.
  • The separation step adds negligible overhead, so the method scales to datasets that are too large for full eigendecomposition while producing test-set embeddings by one forward pass.
  • The proposed Grassmann Score, comparing the leading Laplacian eigenspaces of the input and embedding graphs, distinguishes global-structure-preserving embeddings that kNN and Silhouette scores cannot tell apart.

Reading between the lines

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

  • The same diagonalization trick could be applied to any neural network that minimizes a Rayleigh-quotient loss under an orthogonality constraint, not only SpectralNet; the paper does not explore this generalization.
  • Because the minibatch average is a stochastic estimator of Q^T Λ Q, increasing the number of minibatches T or using importance-sampled batches should improve eigenvector accuracy; the paper does not quantify this trade-off.
  • Grassmann Score could plausibly be used as a training objective or early-stopping criterion for generalizable embedding methods, though the paper presents it only as an evaluation metric.
  • If a theoretical convergence bound for the minibatch average is established, this post-processing approach could replace full eigendecomposition in non-neural spectral settings as well, whenever scalability is the bottleneck.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

1 major / 5 minor

Summary. The paper proposes Sep-SpectralNet, an extension of SpectralNet that adds a post-processing step intended to recover the individual eigenvectors of a graph Laplacian from the rotation-ambiguous output of SpectralNet. The step diagonalizes a minibatch-averaged projected Laplacian, and the authors claim this yields both the eigenvalues and the rotation needed to separate eigenvectors while preserving SpectralNet's scalability and out-of-sample generalization. The paper also introduces NUMAP, a generalizable UMAP variant that uses the separated spectral embedding as an initialization for a second network trained with the UMAP loss, and proposes the Grassmann Score (GS) as a global-structure evaluation metric. Experiments on four real-world datasets report substantially lower sin^2 eigenvector errors for Sep-SpectralNet than for SpectralNet, and NUMAP is reported to improve GS over Parametric UMAP while retaining comparable kNN accuracy.

Significance. If the central claim holds, Sep-SpectralNet would be the first spectral embedding implementation to combine scalability, out-of-sample generalization, and eigenvector separation, and NUMAP would address a known limitation of Parametric UMAP by restoring the spectral initialization that is critical for global structure preservation. The empirical design has real strengths: ground-truth eigenvectors are computed externally on combined train+test graphs, comparisons are run over multiple seeds and datasets, and the improvements in eigenvector separation are large and consistent. The code is publicly available, which supports reproducibility. However, two load-bearing theoretical points need work before the claims can be accepted as stated: the minibatch-averaging step in Algorithm 2 lacks a consistency or bias analysis, and the proof of Lemma 1 in Appendix A is invalid as written. The empirical evidence is suggestive but does not by itself substitute for the missing justification of the core separation step.

major comments (1)
  1. [Sec. 1 and Sec. 4.1] The paper states that SpectralNet 'cannot directly be adapted for SE' and attributes this to the rotation and reflection ambiguity of the Rayleigh quotient, supported by Lemma 1. Given that the proof of Lemma 1 is invalid as written and that the practical separation step rests on the unverified minibatch-averaging approximation, the theoretical claim is currently stronger than what the manuscript establishes. Please align the wording of these claims with the actual level of theoretical support, or supply the missing arguments.
minor comments (5)
  1. [Table 8] The KMNIST v2 entry reads '0.0.044'; this should be '0.044'.
  2. [App. C.2] The text says the graph construction is 'similar to the one used by Gomez et al. in UMAP', but the cited reference is McInnes et al. (2018); please correct the attribution or add the intended citation.
  3. [Algorithm 2 and Sec. 4.2] Algorithm 2 computes random-walk Laplacians, while Lemma 1 and the separation derivation are stated for the unnormalized Laplacian; please state explicitly which Laplacian variant the analysis covers, since the eigenvectors differ across variants.
  4. [Sec. 4.2] The sentence 'The eigenvectors matrix of tilde-Lambda is the inverse of the orthogonal matrix Q' would be clearer as 'the transpose', since diagonalizing tilde-Lambda returns Q^T rather than Q^{-1} in the usual convention.
  5. [Sec. 5.1] Eigenvalue approximation results are reported only for Parkinsons Telemonitoring and Appliances Energy Prediction; please report the eigenvalue correlations for CIFAR10 and KMNIST as well, or explain why they are omitted.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the proposed rotation-recovery step is an exact linear-algebra identity given the eigenspace approximation, and all benchmark comparisons use external ground-truth SVD embeddings.

full rationale

The paper's central derivation (Sec. 4.2, Algorithm 2) rests on the identity (VQ)^T L (VQ) = Q^T Λ Q. Given Lemma 1 — any minimizer of the Rayleigh quotient under orthogonality has the form VQ — diagonalizing the projected Laplacian indeed recovers Q and Λ. This is a mathematical identity, not a fitted parameter renamed as a prediction. Algorithm 2 estimates Λ-tilde from random minibatches; while the paper provides no concentration or bias bound for this estimate, an unproven approximation is a soundness/correctness risk, not circularity, because the target quantity (the eigendecomposition of the full Laplacian) is defined independently of the minibatch estimate. The evaluation in Sec. 5.1 compares against ground-truth eigenvectors computed by SVD/LOBPCG on the train-plus-test affinity matrix, which is an external benchmark. The Grassmann Score is an evaluation metric, and NUMAP is trained with the UMAP loss rather than to minimize GS, so its use in evaluation is not definitionally forced. Lemma 1 is proved in App. A using standard interlacing/eigendecomposition facts and does not invoke a uniqueness theorem imported from the authors' prior work. The only author-overlap citation is SpectralNet (Shaham et al. 2018), used as a building block for the eigenspace approximation; that method is code-reproduced, externally validated, and also empirically re-verified in the present paper's Fig. 2a, so the citation constitutes independent support rather than a circular chain.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The method relies on SpectralNet's empirical convergence, an unproved minibatch averaging heuristic, and standard manifold assumptions. No new physical entities are posited. The only new construct is the Grassmann Score evaluation metric, which is a measurement tool rather than a postulated entity.

free parameters (4)
  • Graph construction n_neighbors = 5-20 (per dataset)
    Number of nearest neighbors used to build the affinity matrix; chosen per dataset (App. C.2, Table 12). Standard hyperparameter, not fitted to target eigenvectors.
  • NUMAP SE dimension k = varies: Cifar10=10, Appliances=5, Wine=10, Banknote=3, MNIST=10, FashionMNIST=10
    Dimension of the spectral embedding fed into the UMAP loss; chosen per dataset (App. G).
  • GS hyperparameter t = 2
    Number of eigenvectors used in the Grassmann Score; set to 2 by default (Sec. 4.5).
  • Batch size m = 512 or 2048
    Minibatch size for SpectralNet training and for the separation averaging; standard training hyperparameter (Table 12).
assumptions (5)
  • domain assumption SpectralNet converges to the eigenspace of the Laplace-Beltrami operator of the underlying manifold
    Taken from Shaham et al. (2018); the separation step presupposes the learned subspace is a good approximation of the true eigenspace.
  • ad hoc to paper The minibatch-averaged matrix (1/T) \sum_i Y_i^T L_i Y_i approximates Q^T \Lambda Q
    Algorithm 2 lines 6-7; no convergence or error bound is provided.
  • standard math Uniqueness of eigendecomposition used in Lemma 1's proof (i.e., eigenvalues are distinct)
    App. A; the proof invokes uniqueness, which holds for simple eigenvalues; with repeated eigenvalues the conclusion still holds but the proof is incomplete.
  • domain assumption UMAP-style graph construction captures the local manifold topology
    Used in both Sep-SpectralNet and GS; standard in UMAP literature (McInnes et al., 2018).
  • domain assumption SE initialization is crucial for UMAP's global structure preservation
    Borrowed from Kobak & Linderman (2021); motivates NUMAP.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Generalizable Spectral Embedding with an Application to UMAP." pith.science (2026). https://pith.science/paper/Q4LZMAGP

@misc{pith2026250111305,
  author       = {Pith},
  title        = {Pith review of: Generalizable Spectral Embedding with an Application to UMAP},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/Q4LZMAGP}},
  note         = {Machine review of arXiv:2501.11305}
}
read the original abstract

Spectral Embedding (SE) is a popular method for dimensionality reduction, applicable across diverse domains. Nevertheless, its current implementations face three prominent drawbacks which curtail its broader applicability: generalizability (i.e., out-of-sample extension), scalability, and eigenvectors separation. Existing SE implementations often address two of these drawbacks; however, they fall short in addressing the remaining one. In this paper, we introduce Sep-SpectralNet (eigenvector-separated SpectralNet), a SE implementation designed to address all three limitations. Sep-SpectralNet extends SpectralNet with an efficient post-processing step to achieve eigenvectors separation, while ensuring both generalizability and scalability. This method expands the applicability of SE to a wider range of tasks and can enhance its performance in existing applications. We empirically demonstrate Sep-SpectralNet's ability to consistently approximate and generalize SE, while maintaining SpectralNet's scalability. Additionally, we show how Sep-SpectralNet can be leveraged to enable generalizable UMAP visualization. Our codes are publicly available.

Figures

Figures reproduced from arXiv: 2501.11305 by the authors.

Figure 1
Figure 1. Spectral Embedding is crucial for global structure preservation. (a) Non-parametric UMAP preserves the global structure when initialized with SE, but fails to do so with PCA initialization - note the in-cluster separation in the middle row; (b) A similar effect is observed in generalizable implementations: P. UMAP, which does not involve SE, , fails to maintain global structure, resulting in overlapping clusters (e.… view at source ↗
Figure 2
Figure 2. (a) Sep-SpectralNet separates the eigenvectors. Approximation of the 2-dimensional SE of the moon dataset (a 2D moon linearly embedded into 10D input space) using SpectralNet (in blue) and Sep￾SpectralNet (in green) over 100 runs, on train (top row) and test (bottom row). Left column: distribution of the Grassmann distance between the output and true subspace. Second to Fourth columns: distribution of the sin2 dista… view at source ↗
Figure 3
Figure 3. Incorporating Sep-SpectralNet for generalizable UMAP. UMAP vs. NUMAP vs. P. UMAP overview. NUMAP integrates SE, as in UMAP, while enabling generalization. First, it learns a parametric representation of the k-dimensional SE. Then, it learns a map from the SE to the final ℓ￾dimensional space, incorporating residual connections from the first ℓ input components to the output layer. both local and global structures eff… view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: Grassmann Score (GS) reflects global structure preservation. A demonstration of the alignment between intuitive expectations and GS on a toy dataset of two 3D tangent spheres. Four possible 2D embeddings of the dataset are shown, alongside their corresponding GS, kNN a…
Figure 5
Figure 5. Figure 5: Sep-SpectralNet successfully separates the eigenvectors. A comparison between Sep￾SpectralNet and SpectralNet of SE and Fiedler Vector (v2) approximation on real-world datasets. The values are the mean and standard deviation of the sin2 distance between the predicted a…
Figure 6
Figure 6. Figure 6: NUMAP and P. UMAP vi￾sualization of the Banknote dataset test set. NUMAP’s enhanced ability to preserve global structure is evident in the clearer separation between clusters com￾pared to P. UMAP. Global Structure. Tab. 3 presents the GS results on these datasets. The …
Figure 7
Figure 7. Figure 7: Time-series data visualization using NUMAP. A comparison between UMAP and NUMAP visualizations of a dynamical system. In the UMAP visualization, the structure and cluster positions vary significantly across timesteps, hindering the interpretability of the underlying pr…
Figure 8
Figure 8. Figure 8: Additional demonstrations of the alignment between the intuitive expectation and the GS results [PITH_FULL_IMAGE:figures/full_fig_p020_8.png]
Figure 9
Figure 9. Figure 9: The catastrophic forgetting phenomenon when fine-tuning Sep-SpectralNet to match UMAP per [PITH_FULL_IMAGE:figures/full_fig_p022_9.png]
Figure 10
Figure 10. Figure 10: Test set Visualizations of NUMAP, NUMAP-SN and P. UMAP on the datasets corresponding to [PITH_FULL_IMAGE:figures/full_fig_p023_10.png]
Figure 11
Figure 11. Figure 11: Fig. 7 extension with P. UMAP [PITH_FULL_IMAGE:figures/full_fig_p024_11.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

69 extracted references · 35 canonical work pages

  1. [1]

    Stefan Aeberhard and M. Forina. Wine . UCI Machine Learning Repository, 1992. DOI : https://doi.org/10.24432/C5PC7J

    doi:10.24432/c5pc7j 1992

  2. [2]

    Local graph partitioning using pagerank vectors

    Reid Andersen, Fan Chung, and Kevin Lang. Local graph partitioning using pagerank vectors. In 2006 47th Annual IEEE Symposium on Foundations of Computer Science (FOCS'06), pp.\ 475--486. IEEE, 2006. doi:10.1109/FOCS.2006.44

    doi:10.1109/focs.2006.44 2006

  3. [3]

    A spectral algorithm for envelope reduction of sparse matrices

    Stephen T Barnard, Alex Pothen, and Horst D Simon. A spectral algorithm for envelope reduction of sparse matrices. In Proceedings of the 1993 ACM/IEEE Conference on Supercomputing, pp.\ 493--502, 1993. doi:10.1145/169627.169790

    work page arXiv 1993

  4. [4]

    Directional graph networks

    Dominique Beaini, Saro Passaro, Vincent L \'e tourneau, Will Hamilton, Gabriele Corso, and Pietro Li \`o . Directional graph networks. In International Conference on Machine Learning, pp.\ 748--758. PMLR, 2021. doi:https://doi.org/10.48550/arXiv.2010.02863

    arXiv doi:10.48550/arxiv.2010.02863 2021

  5. [5]

    Laplacian eigenmaps for dimensionality reduction and data representation

    Mikhail Belkin and Partha Niyogi. Laplacian eigenmaps for dimensionality reduction and data representation. Neural computation, 15 0 (6): 0 1373--1396, 2003. doi:10.1162/089976603321780317

    doi:10.1162/089976603321780317 2003

  6. [6]

    Convergence of laplacian eigenmaps

    Mikhail Belkin and Partha Niyogi. Convergence of laplacian eigenmaps. Advances in neural information processing systems, 19, 2006. doi:https://doi.org/10.7551/mitpress/7503.003.0021

    doi:10.7551/mitpress/7503.003.0021 2006

  7. [7]

    Locally optimal block preconditioned conjugate gradient method for hierarchical matrices

    Peter Benner and Thomas Mach. Locally optimal block preconditioned conjugate gradient method for hierarchical matrices. PAMM, 11 0 (1): 0 741--742, 2011. doi:https://doi.org/10.1002/pamm.201110360

    work page doi:10.1002/pamm.201110360 2011

  8. [8]

    Algebraic multigrid (amg) for sparse matrix equations

    Achi Brandt, Steve McCormick, and John Ruge. Algebraic multigrid (amg) for sparse matrix equations. Sparsity and its Applications, 257, 1984

    work page 1984

Show all 69 references
  1. [9]

    Dynamic graph theoretical analysis of functional connectivity in parkinson's disease: The importance of fiedler value

    Jiayue Cai, Aiping Liu, Taomian Mi, Saurabh Garg, Wade Trappe, Martin J McKeown, and Z Jane Wang. Dynamic graph theoretical analysis of functional connectivity in parkinson's disease: The importance of fiedler value. IEEE journal of biomedical and health informatics, 23 0 (4):...

    2018

  2. [10]

    Laplacian eigenmaps and principal curves for high resolution pseudotemporal ordering of single-cell rna-seq profiles

    Kieran Campbell, Chris P Ponting, and Caleb Webber. Laplacian eigenmaps and principal curves for high resolution pseudotemporal ordering of single-cell rna-seq profiles. bioRxiv, pp.\ 027219, 2015. doi:https://doi.org/10.1101/027219

    2015 doi

  3. [11]

    Appliances Energy Prediction

    Luis Candanedo. Appliances Energy Prediction . UCI Machine Learning Repository, 2017. DOI : https://doi.org/10.24432/C5VC8G

    2017 doi

  4. [12]

    Specnet2: Orthogonalization-free spectral embedding by neural networks

    Ziyu Chen, Yingzhou Li, and Xiuyuan Cheng. Specnet2: Orthogonalization-free spectral embedding by neural networks. arXiv preprint arXiv:2206.06644, 2022. doi:https://doi.org/10.48550/arXiv.2206.06644

    2022 doi arXiv

  5. [13]

    Intrinsic map dynamics exploration for uncharted effective free-energy landscapes

    Eliodoro Chiavazzo, Roberto Covino, Ronald R Coifman, C William Gear, Anastasia S Georgiou, Gerhard Hummer, and Ioannis G Kevrekidis. Intrinsic map dynamics exploration for uncharted effective free-energy landscapes. Proceedings of the National Academy of Sciences, 114 0 (28):...

    2017 doi

  6. [14]

    Deep learning for classical japanese literature, 2018

    Tarin Clanuwat, Mikel Bober-Irizar, Asanobu Kitamoto, Alex Lamb, Kazuaki Yamamoto, and David Ha. Deep learning for classical japanese literature, 2018

    2018

  7. [15]

    Geometric harmonics: a novel tool for multiscale out-of-sample extension of empirical functions

    Ronald R Coifman and St \'e phane Lafon. Geometric harmonics: a novel tool for multiscale out-of-sample extension of empirical functions. Applied and Computational Harmonic Analysis, 21 0 (1): 0 31--52, 2006 a . doi:https://doi.org/10.1016/j.acha.2005.07.005

    2006 doi

  8. [16]

    Diffusion maps

    Ronald R Coifman and St \'e phane Lafon. Diffusion maps. Applied and computational harmonic analysis, 21 0 (1): 0 5--30, 2006 b . doi:https://doi.org/10.1016/j.acha.2006.04.006

    2006 doi

  9. [17]

    From t -sne to umap with contrastive learning

    Sebastian Damrich, Jan Niklas B \"o hm, Fred A Hamprecht, and Dmitry Kobak. From t -sne to umap with contrastive learning. arXiv preprint arXiv:2206.01816, 2022. doi:https://doi.org/10.48550/arXiv.2206.01816

    2022 doi arXiv

  10. [18]

    Convolutional neural networks on graphs with fast localized spectral filtering

    Micha \"e l Defferrard, Xavier Bresson, and Pierre Vandergheynst. Convolutional neural networks on graphs with fast localized spectral filtering. Advances in neural information processing systems, 29, 2016. doi:https://doi.org/10.48550/arXiv.1606.09375

    2016 doi arXiv

  11. [19]

    The mnist database of handwritten digit images for machine learning research [best of the web]

    Li Deng. The mnist database of handwritten digit images for machine learning research [best of the web]. IEEE Signal Processing Magazine, 29 0 (6): 0 141--142, 2012. doi:10.1109/MSP.2012.2211477

    2012

  12. [20]

    Neuralef: Deconstructing kernels by deep neural networks

    Zhijie Deng, Jiaxin Shi, and Jun Zhu. Neuralef: Deconstructing kernels by deep neural networks. In International Conference on Machine Learning, pp.\ 4976--4992. PMLR, 2022. doi:https://doi.org/10.48550/arXiv.2205.00165

    2022 doi arXiv

  13. [21]

    Compression of cerebellar functional gradients in schizophrenia

    Debo Dong, Cheng Luo, Xavier Guell, Yulin Wang, Hui He, Mingjun Duan, Simon B Eickhoff, and Dezhong Yao. Compression of cerebellar functional gradients in schizophrenia. Schizophrenia bulletin, 46 0 (5): 0 1282--1295, 2020. doi:10.1093/schbul/sbaa016

    2020 doi

  14. [22]

    Visual exploration of relationships and structure in low-dimensional embeddings

    Klaus Eckelt, Andreas Hinterreiter, Patrick Adelberger, Conny Walchshofer, Vaishali Dhanoa, Christina Humer, Moritz Heckmann, Christian Steinparz, and Marc Streit. Visual exploration of relationships and structure in low-dimensional embeddings. IEEE Transactions on Visualizati...

    2023

  15. [23]

    Algebraic connectivity of graphs

    Miroslav Fiedler. Algebraic connectivity of graphs. Czechoslovak mathematical journal, 23 0 (2): 0 298--305, 1973. doi:10.21136/CMJ.1973.101168

    1973

  16. [24]

    A property of eigenvectors of nonnegative symmetric matrices and its application to graph theory

    Miroslav Fiedler. A property of eigenvectors of nonnegative symmetric matrices and its application to graph theory. Czechoslovak mathematical journal, 25 0 (4): 0 619--633, 1975. doi:10.21136/CMJ.1975.101357

    1975

  17. [25]

    Eigengame: Pca as a nash equilibrium

    Ian Gemp, Brian McWilliams, Claire Vernade, and Thore Graepel. Eigengame: Pca as a nash equilibrium. arXiv preprint arXiv:2010.00554, 2020. doi:https://doi.org/10.48550/arXiv.2010.00554

    2010 doi arXiv

  18. [26]

    Sampling a rare protein transition using quantum annealing

    Danial Ghamari, Roberto Covino, and Pietro Faccioli. Sampling a rare protein transition using quantum annealing. Journal of Chemical Theory and Computation, 20 0 (8): 0 3322--3334, 2024. doi:https://doi.org/10.1021/acs.jctc.3c01174

    2024 doi

  19. [27]

    Similarity search in high dimensions via hashing

    Aristides Gionis, Piotr Indyk, Rajeev Motwani, et al. Similarity search in high dimensions via hashing. In Vldb, volume 99, pp.\ 518--529, 1999

    1999

  20. [28]

    Unsupervised learning methods for molecular simulation data

    Aldo Glielmo, Brooke E Husic, Alex Rodriguez, Cecilia Clementi, Frank No \'e , and Alessandro Laio. Unsupervised learning methods for molecular simulation data. Chemical Reviews, 121 0 (16): 0 9722--9758, 2021. doi:https://doi.org/10.1021/acs.chemrev.0c01195

    2021 doi

  21. [29]

    Revealing the hidden structure of disordered materials by parameterizing their local structural manifold

    Thomas J Hardin, Michael Chandross, Rahul Meena, Spencer Fajardo, Dimitris Giovanis, Ioannis Kevrekidis, Michael L Falk, and Michael D Shields. Revealing the hidden structure of disordered materials by parameterizing their local structural manifold. Nature communications, 15 0...

    2024 doi

  22. [30]

    Alternating diffusion maps for multimodal data fusion

    Ori Katz, Ronen Talmon, Yu-Lun Lo, and Hau-Tieng Wu. Alternating diffusion maps for multimodal data fusion. Information Fusion, 45: 0 346--360, 2019. doi:https://doi.org/10.1016/j.inffus.2018.01.007

    2019 doi

  23. [31]

    Parametric t-stochastic neighbor embedding with quantum neural network

    Yoshiaki Kawase, Kosuke Mitarai, and Keisuke Fujii. Parametric t-stochastic neighbor embedding with quantum neural network. Physical Review Research, 4 0 (4): 0 043199, 2022. doi:https://doi.org/10.48550/arXiv.2202.04238

    2022 doi arXiv

  24. [32]

    A survey of machine learning techniques applied to self-organizing cellular networks

    Paulo Valente Klaine, Muhammad Ali Imran, Oluwakayode Onireti, and Richard Demo Souza. A survey of machine learning techniques applied to self-organizing cellular networks. IEEE Communications Surveys & Tutorials, 19 0 (4): 0 2392--2431, 2017. doi:10.1109/COMST.2017.2727878

    2017

  25. [33]

    Initialization is critical for preserving global data structure in both t-sne and umap

    Dmitry Kobak and George C Linderman. Initialization is critical for preserving global data structure in both t-sne and umap. Nature biotechnology, 39 0 (2): 0 156--157, 2021. doi:https://doi.org/10.1038/nbt.4314

    2021 doi

  26. [34]

    Automatic domain decomposition of proteins by a gaussian network model

    Sibsankar Kundu, Dan C Sorensen, and George N Phillips Jr. Automatic domain decomposition of proteins by a gaussian network model. Proteins: Structure, Function, and Bioinformatics, 57 0 (4): 0 725--733, 2004

    2004

  27. [35]

    Data fusion and multicue data matching by diffusion maps

    Stephane Lafon, Yosi Keller, and Ronald R Coifman. Data fusion and multicue data matching by diffusion maps. IEEE Transactions on pattern analysis and machine intelligence, 28 0 (11): 0 1784--1797, 2006. doi:10.1109/TPAMI.2006.223

    2006 doi

  28. [36]

    Learning the geometry of common latent variables using alternating-diffusion

    Roy R Lederman and Ronen Talmon. Learning the geometry of common latent variables using alternating-diffusion. Applied and Computational Harmonic Analysis, 44 0 (3): 0 509--536, 2018. doi:https://doi.org/10.1016/j.acha.2015.09.002

    2018 doi

  29. [37]

    ARPACK users' guide: solution of large-scale eigenvalue problems with implicitly restarted Arnoldi methods

    Richard B Lehoucq, Danny C Sorensen, and Chao Yang. ARPACK users' guide: solution of large-scale eigenvalue problems with implicitly restarted Arnoldi methods. SIAM, 1998

    1998

  30. [38]

    Data classification methodology for electronic noses using uniform manifold approximation and projection and extreme learning machine

    Jersson X Leon-Medina, N \'u ria Par \'e s, Maribel Anaya, Diego A Tibaduiza, and Francesc Pozo. Data classification methodology for electronic noses using uniform manifold approximation and projection and extreme learning machine. Mathematics, 10 0 (1): 0 29, 2021. doi:https:...

    2021 doi

  31. [39]

    Rayleigh quotient based optimization methods for eigenvalue problems

    Ren-Cang Li. Rayleigh quotient based optimization methods for eigenvalue problems. In Matrix Functions and Matrix Equations, pp.\ 76--108. World Scientific, 2015. doi:https://doi.org/10.1142/9789814675772_0004

    2015 doi

  32. [40]

    Sign and basis invariant networks for spectral graph representation learning

    Derek Lim, Joshua Robinson, Lingxiao Zhao, Tess Smidt, Suvrit Sra, Haggai Maron, and Stefanie Jegelka. Sign and basis invariant networks for spectral graph representation learning. arXiv preprint arXiv:2202.13013, 2022. doi:https://doi.org/10.48550/arXiv.2202.13013

    2022 doi arXiv

  33. [41]

    Umap-pytorch: Umap (uniform manifold approximation and projection) in pytorch, 2024

    Leland Zhang Liu. Umap-pytorch: Umap (uniform manifold approximation and projection) in pytorch, 2024. URL https://github.com/elyxlz/umap_pytorch

    2024

  34. [42]

    Banknote Authentication

    Volker Lohweg. Banknote Authentication . UCI Machine Learning Repository, 2012. DOI : https://doi.org/10.24432/C55P57

    2012 doi

  35. [43]

    Laplacian canonization: A minimalist approach to sign and basis invariant spectral embedding

    George Ma, Yifei Wang, and Yisen Wang. Laplacian canonization: A minimalist approach to sign and basis invariant spectral embedding. Advances in Neural Information Processing Systems, 36, 2024. doi:https://doi.org/10.48550/arXiv.2310.18716

    2024 doi arXiv

  36. [44]

    Umap: Uniform manifold approximation and projection for dimension reduction

    Leland McInnes, John Healy, and James Melville. Umap: Uniform manifold approximation and projection for dimension reduction. arXiv preprint arXiv:1802.03426, 2018. doi:https://doi.org/10.48550/arXiv.1802.03426

    2018 doi arXiv

  37. [45]

    Diffusion nets

    Gal Mishne, Uri Shaham, Alexander Cloninger, and Israel Cohen. Diffusion nets. Applied and Computational Harmonic Analysis, 47 0 (2): 0 259--285, 2019. doi:https://doi.org/10.48550/arXiv.1506.07840

    2019 doi arXiv

  38. [46]

    o m. \"U ber die praktische aufl \

    Evert J Nystr \"o m. \"U ber die praktische aufl \"o sung von integralgleichungen mit anwendungen auf randwertaufgaben. 1930. doi:https://doi.org/10.1007/BF02547521

    1930 doi

  39. [47]

    Graph signal processing: Overview, challenges, and applications

    Antonio Ortega, Pascal Frossard, Jelena Kova c evi \'c , Jos \'e MF Moura, and Pierre Vandergheynst. Graph signal processing: Overview, challenges, and applications. Proceedings of the IEEE, 106 0 (5): 0 808--828, 2018. doi:10.1109/JPROC.2018.2820126

    2018

  40. [48]

    Differences in subcortico-cortical interactions identified from connectome and microcircuit models in autism

    Bo-yong Park, Seok-Jun Hong, Sofie L Valk, Casey Paquola, Oualid Benkarim, Richard AI Bethlehem, Adriana Di Martino, Michael P Milham, Alessandro Gozzi, BT Thomas Yeo, et al. Differences in subcortico-cortical interactions identified from connectome and microcircuit models in ...

    2021 doi

  41. [49]

    Multiscale neural gradients reflect transdiagnostic effects of major psychiatric conditions on cortical morphology

    Bo-yong Park, Valeria Kebets, Sara Larivi \`e re, Meike D Hettwer, Casey Paquola, Daan van Rooij, Jan Buitelaar, Barbara Franke, Martine Hoogman, Lianne Schmaal, et al. Multiscale neural gradients reflect transdiagnostic effects of major psychiatric conditions on cortical morp...

    2022 doi

  42. [50]

    Spectral inference networks: Unifying deep and spectral learning

    David Pfau, Stig Petersen, Ashish Agarwal, David GT Barrett, and Kimberly L Stachenfeld. Spectral inference networks: Unifying deep and spectral learning. arXiv preprint arXiv:1806.02215, 2018. doi:https://doi.org/10.48550/arXiv.1806.02215

    2018 doi arXiv

  43. [51]

    Parametric umap embeddings for representation and semisupervised learning

    Tim Sainburg, Leland McInnes, and Timothy Q Gentner. Parametric umap embeddings for representation and semisupervised learning. Neural Computation, 33 0 (11): 0 2881--2907, 2021. doi:https://doi.org/10.48550/arXiv.2009.12981

    2021 doi arXiv

  44. [52]

    Using genetic programming to find functional mappings for umap embeddings

    Finn Schofield and Andrew Lensen. Using genetic programming to find functional mappings for umap embeddings. In 2021 IEEE Congress on Evolutionary Computation (CEC), pp.\ 704--711. IEEE, 2021. doi:10.1109/CEC45853.2021.9504848

    2021

  45. [53]

    Spectralnet: Spectral clustering using deep neural networks

    Uri Shaham, Kelly Stanton, Henri Li, Boaz Nadler, Ronen Basri, and Yuval Kluger. Spectralnet: Spectral clustering using deep neural networks. In Proc. ICLR 2018, 2018. doi:https://doi.org/10.48550/arXiv.1801.01587

    2018 doi arXiv

  46. [54]

    Amino acid partitioning using a fiedler vector model

    SJ Shepherd, CB Beggs, and Sue Jones. Amino acid partitioning using a fiedler vector model. European Biophysics Journal, 37: 0 105--109, 2007. doi:https://doi.org/10.1007/s00249-007-0182-y

    2007 doi

  47. [55]

    Convergence of laplacian spectra from random samples

    Zuoqiang Shi. Convergence of laplacian spectra from random samples. arXiv preprint arXiv:1507.00151, 2015. doi:https://doi.org/10.48550/arXiv.1507.00151

    2015 doi arXiv

  48. [56]

    Basis: batch aligned spectral embedding space

    Or Streicher, Ido Cohen, and Guy Gilboa. Basis: batch aligned spectral embedding space. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp.\ 10396--10405, 2023. doi:https://doi.org/10.48550/arXiv.2211.16960

    2023 doi arXiv

  49. [57]

    Fiedler regularization: Learning neural networks with graph sparsity

    Edric Tam and David Dunson. Fiedler regularization: Learning neural networks with graph sparsity. In International Conference on Machine Learning, pp.\ 9346--9355. PMLR, 2020. doi:https://doi.org/10.48550/arXiv.2003.00992

    2020 doi arXiv

  50. [58]

    Parkinsons Telemonitoring

    Athanasios Tsanas and Max Little. Parkinsons Telemonitoring . UCI Machine Learning Repository, 2009. DOI : https://doi.org/10.24432/C5ZS3N

    2009 doi

  51. [59]

    Learning a parametric embedding by preserving local structure

    Laurens Van Der Maaten. Learning a parametric embedding by preserving local structure. In Artificial intelligence and statistics, pp.\ 384--391. PMLR, 2009

    2009

  52. [60]

    A tutorial on spectral clustering

    Ulrike Von Luxburg. A tutorial on spectral clustering. Statistics and computing, 17: 0 395--416, 2007. doi:https://doi.org/10.48550/arXiv.0711.0189

    2007 doi arXiv

  53. [61]

    Fashion-mnist: a novel image dataset for benchmarking machine learning algorithms

    Han Xiao, Kashif Rasul, and Roland Vollgraf. Fashion-mnist: a novel image dataset for benchmarking machine learning algorithms. arXiv preprint arXiv:1708.07747, 2017. doi:10.48550/arXiv.1708.07747

    2017 doi arXiv

  54. [62]

    Integrative multiscale biochemical mapping of the brain via deep-learning-enhanced high-throughput mass spectrometry

    Yuxuan Richard Xie, Daniel C Castro, Stanislav S Rubakhin, Timothy J Trinklein, Jonathan V Sweedler, and Fan Lam. Integrative multiscale biochemical mapping of the brain via deep-learning-enhanced high-throughput mass spectrometry. Nature Methods, 21: 0 521--530, 2024. doi:htt...

    2024 doi

  55. [63]

    Robust parametric umap for the analysis of single-cell data

    Bingxian Xu and Guangzheng Zhang. Robust parametric umap for the analysis of single-cell data. bioRxiv, pp.\ 2023--11, 2023. doi:https://doi.org/10.1101/2023.11.14.567092

    2023 doi

  56. [64]

    Online learning of open-set speaker identification by active user-registration

    Eunkyung Yoo, Hyeonseop Song, Taehyeong Kim, and Chul Lee. Online learning of open-set speaker identification by active user-registration. In INTERSPEECH, pp.\ 5065--5069, 2022

    2022

  57. [65]

    Eigen-gnn: A graph structure preserving plug-in for gnns

    Ziwei Zhang, Peng Cui, Jian Pei, Xin Wang, and Wenwu Zhu. Eigen-gnn: A graph structure preserving plug-in for gnns. IEEE Transactions on Knowledge and Data Engineering, 35 0 (3): 0 2544--2555, 2021. doi:https://doi.org/10.48550/arXiv.2006.04330

    2021 doi arXiv

  58. [66]

    Delineation of folding pathways of a -sheet miniprotein

    Wenwei Zheng, Bo Qi, Mary A Rohrdanz, Amedeo Caflisch, Aaron R Dinner, and Cecilia Clementi. Delineation of folding pathways of a -sheet miniprotein. The Journal of Physical Chemistry B, 115 0 (44): 0 13065--13074, 2011. doi:https://doi.org/10.1021/jp2076935

    2011 doi

  59. [67]

    Rapid exploration of configuration space with diffusion-map-directed molecular dynamics

    Wenwei Zheng, Mary A Rohrdanz, and Cecilia Clementi. Rapid exploration of configuration space with diffusion-map-directed molecular dynamics. The journal of physical chemistry B, 117 0 (42): 0 12769--12776, 2013. doi:https://doi.org/10.1021/jp401911h

    2013 doi

  60. [68]

    A fiedler vector scoring approach for novel rna motif selection

    Qiyao Zhu and Tamar Schlick. A fiedler vector scoring approach for novel rna motif selection. The Journal of Physical Chemistry B, 125 0 (4): 0 1144--1155, 2021. doi:https://doi.org/10.1021/acs.jpcb.0c10685

    2021 doi

  61. [69]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.