betaFeedback

WorksMathematics

Model-Based Learning of Latent Euclidean Embedding from Relational Data

Abstract

Visualizing and interpreting the dependence structure in relational data often benefits from geometric representations via low-dimensional Euclidean embeddings. However, existing approaches are typically restricted to specific response types, rely on burdensome high-dimensional nonconvex optimization, or lack uniform theoretical guarantees for simultaneous coordinate recovery. To address these limitations, we introduce a unified, distance-based latent space framework for both unipartite and bipartite relational data—accommodating continuous, binary, ordinal, and count-valued observations—via a modular two-stage procedure that decouples signal estimation from geometric recovery. First, we estimate the low-rank signal matrix using spectral denoising. For models with nonlinear link functions, we further apply one-step row- and column-wise refinement to achieve uniform entrywise accuracy. Second, we recover the latent configurations through direct spectral analysis for unipartite data and a low-dimensional alignment optimization for bipartite data. We establish uniform recovery guarantees that characterize the propagation of error from signal estimation to the embedding space. To illustrate the scalability and interpretability of our approach, we apply the proposed method to two real-world datasets: a statistician citation network and European Parliament voting records.

Introduction

Relational data arise in a wide range of scientific and machine-learning applications, including social and information networks, recommender systems, voting records, citation patterns, and user–item interactions. Such data record associations either among objects from a single set (unipartite relational data, such as networks) or between objects from two distinct sets (bipartite relational data, such as user–item interactions). These two settings are geometrically connected. After pooling the two sets, a bipartite relation can be represented as a symmetric relation matrix in which the cross-set blocks are observed, whereas the two within-set blocks are structurally missing.

A central task is to uncover the relational structure among these objects for prediction, scientific interpretation, and exploratory visualization. A broad range of low-dimensional representation methods has been developed for this purpose. Multidimensional scaling (MDS) represents one set of objects from pairwise dissimilarities [38, 7], whereas multidimensional unfolding (MDU) jointly represents two sets of objects from a rectangular matrix of cross-set relations or preferences [35, 7]. Latent space models assign embedding vectors to network nodes and relate the observed network to these embedding vectors [24] (Handcock et al. 2007). Correspondence analysis (CA) and biplots provide joint low-dimensional displays of the rows and columns of a data table [17, 18], while factor models represent two sets of objects through latent factor scores and loading vectors [3, 36, 5]. More broadly, machine learning has developed a diverse family of low-dimensional embedding methods for visualization and representation learning, including manifold and spectral methods, neighbor embeddings, and neural parametric approaches [37, 6, 22, 39, 34 de Bodt] et al., 2025). Although all these methods produce low-dimensional coordinates, proximity in the resulting representation does not have a universal interpretation. Depending on the method, the coordinates are designed to represent pairwise distances, inner-product interactions, standardized departures from independence, or local neighborhood relations.

Among these representations, a common low-dimensional Euclidean configuration is particularly attractive because it places all objects in a shared metric space, so that Euclidean distance has a direct relational interpretation ([35, 7]). Depending on the application, proximity may encode similarity, preference, approval, affinity, or interaction intensity. The resulting geometry can reveal clusters, latent continua, polarization, and objects occupying intermediate positions. Figure 1 provides a schematic illustration of this common Euclidean view for unipartite and bipartite relational data, where the smaller the distance between objects in the embedding space, the stronger their similarity tends to be as reflected by the observed relational data.

Schematic illustrations of latent Euclidean configurations for unipartite and bipartite relational data

Figure 1. Schematic illustrations of latent Euclidean configurations for unipartite (left) and bipartite (right) relational data. In the unipartite setting, relations are observed among objects from a single set; in the bipartite setting, objects from two distinct sets are jointly represented in a common Euclidean space while only cross-set relations are observed.

Despite the utility of low-dimensional representations, existing approaches have four major limitations.

  1. Validity of joint geometric representation. Many methods can provide a joint display of objects in a low-dimensional space, but this does not necessarily provide a valid joint metric geometry. In that case, the interpretation of low-dimensional coordinates depends on the structure that the method is designed to represent. For example, factor models represent row objects by factor scores and column objects by loadings, whose bilinear signal is invariant under paired nonsingular transformations [36, 5]. Biplots and correspondence analysis provide joint displays, but their cross-set interpretation is based on inner products, projections, or standardized associations rather than row–column Euclidean distances [17, 18]. Neighborhood-embedding methods primarily preserve local proximity rather than global Euclidean geometry [14, 39, 34].

  1. Modeling scope. While model-based approaches can yield valid joint geometric interpretations, current methods remain limited to specific observation structures and response types. For instance, multidimensional scaling (MDS), multidimensional unfolding (MDU), and distance-based latent space models are typically tailored to narrow settings [7, 13, 24] (Handcock et al. 2007), whereas more flexible generalized factor models do not induce a common Euclidean metric geometry [36, 5]. Consequently, a unified distance-based framework that handles both unipartite and bipartite relational data across continuous, binary, ordinal, and count-valued responses is currently lacking.

  1. Computational scalability. Classical MDS admits a direct eigendecomposition-based solution [38, 7]. In contrast, metric and nonmetric MDS, classical and probabilistic MDU, and likelihood-based latent space models

commonly estimate one embedding vector per object through a high-dimensional non-convex optimization ([26, 8, 24, 13]). The dimension of such optimization problems grows with the number of objects, making these methods unsuitable for large data sets.

  1. Uniform configuration recovery theory. By uniform configuration recovery, we mean simultaneous control of the estimation errors of all individual embedding vectors, rather than only an average error over the entire configuration. Such guarantees have been studied in selected unipartite settings ([44, 31] (Li et al. 2020)), whereas model-based theory for bipartite Euclidean unfolding—where only cross-set relations are observed—has primarily concerned average configuration recovery under more specific response and identifiability assumptions ([35, 13]). A unified uniform-recovery theory covering both observation structures is lacking.

To address these limitations, we develop a unified distance-based formulation for learning latent Euclidean configurations from both unipartite and bipartite relational data. We assume that the mean of each observation is linked, through a prespecified strictly increasing link function, to a drift parameter minus the squared Euclidean distance between the corresponding embedding vectors. This formulation accommodates continuous, binary, ordinal, and count-valued relations while retaining a common proximity-based interpretation.

We propose a novel estimation procedure that uses a common signal-denoising stage under both relational settings, while adapting the geometric recovery step to their different observation structures. By separating signal estimation from geometric recovery, the procedure avoids joint optimization over all latent positions, and the recovery stage can be combined with any sufficiently accurate signal estimator. We first estimate the signal matrix using a global spectral procedure; for models with nonlinear link functions, we further apply one-step row- and column-wise estimating-equation refinement to obtain uniform entrywise accuracy. Given the denoised signal matrix, the unipartite configuration is recovered by the classical MDS construction based on double centering and eigendecomposition [38, 19]. For bipartite data, the structurally missing within-set blocks prevent this construction from recovering the joint configuration: double centering identifies only the cross-product of the two centered sets of embedding vectors. To tackle this issue, we profile out the high-dimensional singular spaces and formulate this remaining ambiguity as an alignment problem, which is solved by an optimization whose dimension depends only on the latent dimension, rather than on the numbers of row and column objects.

Theoretically, we provide convergence guarantees linking the estimated signal matrix to the recovered Euclidean configuration. For bipartite data, we establish general identifiability conditions combining a quantitative measure of local identification strength with global separation. The resulting error bounds explicitly adapt to the local identification strength and are more general than the anchor-point conditions imposed in previous work [13]. Based on these identifiability conditions, our results further quantify how uniform accuracy of the estimated signal matrix propagates to simultaneous recovery of all embedding vectors. Taken together, these results establish a unified signal-to-configuration recovery theory for both unipartite and bipartite relational data.

Related Literature

Multidimensional scaling (MDS) and multidimensional unfolding (MDU) are most closely related to our work in terms of the target parameter, as both aim to recover a low-dimensional Euclidean configuration from relational or proximity data [38, 35, 7]. Classical and modern MDS/MDU methods recover such configurations through spectral or optimization-based procedures [7] de Leeuw and Mair, 2009). Statistical formulations of unfolding further treat the latent coordinates as model parameters and establish identifiability and configuration recovery guarantees under probabilistic response models [13]. Our work shares the goal of recovering an interpretable Euclidean geometry, but considers a common mean formulation for both unipartite and bipartite data, separates signal estimation from geometric recovery, and develops uniform configuration recovery guarantees.

Distance-based latent space models provide another closely related framework for model-based geometric representation of network data, in which each node is assigned a latent position and edge probabilities depend on distances between latent positions [24] (Handcock et al. 2007). Estimation has traditionally relied on likelihood or Bayesian procedures, while more scalable methods based on nuclear-norm regularization and projected gradient algorithms have been developed for broad classes of latent space models (Ma et al., 2020). Recent work has also established uniform consistency and asymptotic distribution theory for latent-position estimators under a range of edge types and link functions (Li et al., 2025). These developments primarily concern unipartite networks, whereas our framework treats unipartite and bipartite observation structures jointly and, in the latter case, addresses configuration recovery when only cross-set relations are observed.

Our signal-estimation problem is also connected to the literature on high-dimensional factor and generalized latent factor models, which represents a large data matrix through low-dimensional factors and loadings and targets the resulting bilinear low-rank structure [3, 16, 12]. Principal-component, projected-PCA, maximum-likelihood, and related scalable methods have been developed for estimating such latent structures in high dimensions [16, 41, 25]. This literature provides consistency, convergence rates, inferential theory, and results on factor identifiability under various high-dimensional regimes. In our setting, however, the low-rank factorization is primarily an intermediate device for estimating the signal matrix: the final target is a common Euclidean configuration whose row and column coordinates have the same metric interpretation, requiring an additional geometric recovery step beyond factor estimation.

Finally, our signal-estimation procedure is related to a growing literature on spectral estimation and refinement for high-dimensional latent-variable models. SVD-based methods exploit low-rank structure in the population signal to construct computationally efficient estimators or initializers, for example in high-dimensional item factor analysis, mixed-membership models, and deep latent-variable models [43, 28] (Chen and Gu 2024). Related spectral-initialization ideas have also been developed for low-rank matrix and tensor estimation, with statistical and computational guarantees [42]. More recent work combines spectral initialization with model-specific one-step refinement to improve the statistical accuracy of the initial spectral estimator in high-dimensional latent-variable models [32] (Chen and Li 2024). Our procedure follows this broad spectral-plus-refinement principle for models with nonlinear link functions, while its output is a uniformly accurate estimate of a distance-based signal matrix that is subsequently passed to a separate configuration recovery procedure; in the bipartite setting, this additionally requires resolving the non-orthogonal alignment ambiguity left by the cross-product structure.

Notation

We introduce some notation for the rest of the paper. For a positive integer NN, let [N]:={1,…,N}[N] := \{1,\ldots,N\}. For a matrix M\mathbf{M}, let vec⁡(M)\operatorname{vec}(\mathbf{M}) denote its column-wise vectorization. Let SK:={diag⁡(q1,…,qK):qk∈{−1,1}}\mathcal{S}_K := \{\operatorname{diag}(q_1,\ldots,q_K): q_k \in\{-1,1\}\} denote the set of K×KK \times K diagonal sign matrices. For two integers a,ba,b, let a∨b=max⁡{a,b}a \vee b = \max\{a,b\} and a∧b=min⁡{a,b}a \wedge b = \min\{a,b\}. We let ∥x∥\lVert x\rVert denote the standard Euclidean norm of a vector x=(x1,…,xn)T\mathbf{x}=(x_1,\ldots,x_n)^{\mathsf{T}} and let ∥x∥∞=max⁡i∣xi∣\|\mathbf{x}\|_{\infty}=\max_i|x_i| denote the infinity norm of x\mathbf{x}. For a matrix X=(xij)m×n\mathbf{X}=(x_{ij})_{m\times n} and a function f:R→Rf:\mathbb{R}\to\mathbb{R}, let f(X):=(f(xij))m×nf(\mathbf{X}):=(f(x_{ij}))_{m\times n}. The matrix max-norm is denoted by ∥X∥max⁡=max⁡i,j∣xij∣\|\mathbf{X}\|_{\max}=\max_{i,j}|x_{ij}|. In general, we define the operator pp-norm of a matrix by ∥X∥p=sup⁡∥u∥p=1∥Xu∥p\|\mathbf{X}\|_p=\sup_{\|\mathbf{u}\|_p=1}\|\mathbf{X}\mathbf{u}\|_p. In particular, ∥X∥1=max⁡j∑i=1m∣xij∣\|\mathbf{X}\|_1=\max_j\sum_{i=1}^{m}|x_{ij}|; ∥X∥∞=max⁡i∑j=1n∣xij∣\|\mathbf{X}\|_{\infty}=\max_i\sum_{j=1}^{n}|x_{ij}|; and ∥⋅∥\|\cdot\| denotes the spectral norm, or equivalently the matrix 2-norm ∥⋅∥2\|\cdot\|_2, when there is no ambiguity. Let ∥X∥F\|\mathbf{X}\|_F and ∥X∥2→∞\|\mathbf{X}\|_{2\to\infty} denote its Frobenius and two-to-infinity norms, respectively, which are defined as ∥X∥F=(∑i=1m∑j=1n∣xij∣2)1/2\|\mathbf{X}\|_F=\left(\sum_{i=1}^{m}\sum_{j=1}^{n}|x_{ij}|^2\right)^{1/2} and ∥X∥2→∞=max⁡i(∑j=1nxij2)1/2\|\mathbf{X}\|_{2\to\infty}=\max_i\left(\sum_{j=1}^{n}x_{ij}^2\right)^{1/2}. ∥X∥∗\|\mathbf{X}\|_* denotes the nuclear norm of X\mathbf{X}, which is the sum of all singular values. For any matrix A∈RN×J\mathbf{A}\in\mathbb{R}^{N\times J} with singular value decomposition A=∑i=1N∧JσiuiviT\mathbf{A}=\sum_{i=1}^{N\wedge J}\sigma_i\mathbf{u}_i\mathbf{v}_i^{\mathsf{T}} and σ1≥⋯≥σN∧J≥0\sigma_1\ge\cdots\ge\sigma_{N\wedge J}\ge0, we define SVD⁡K(A):=∑i=1KσiuiviT\operatorname{SVD}_K(\mathbf{A}):=\sum_{i=1}^{K}\sigma_i\mathbf{u}_i\mathbf{v}_i^{\mathsf{T}}. For any symmetric matrix M∈RN×N\mathbf{M}\in\mathbb{R}^{N\times N} with eigendecomposition M=∑i=1NλiuiuiT\mathbf{M}=\sum_{i=1}^{N}\lambda_i\mathbf{u}_i\mathbf{u}_i^{\mathsf{T}} and λ1≥⋯≥λN\lambda_1\ge\cdots\ge\lambda_N, we define Eigen⁡K(M):=∑i=1KλiuiuiT\operatorname{Eigen}_K(\mathbf{M}):=\sum_{i=1}^{K}\lambda_i\mathbf{u}_i\mathbf{u}_i^{\mathsf{T}}. For a<ba<b, let P[a,b](x)=a1{x≤a}+x1{a<x<b}+b1{x≥b}\mathcal{P}_{[a,b]}(x)=a\mathbf{1}_{\{x\le a\}}+x\mathbf{1}_{\{a<x<b\}}+b\mathbf{1}_{\{x\ge b\}}. For a matrix M=(mij)\mathbf{M}=(m_{ij}), let P[a,b](M)\mathcal{P}_{[a,b]}(\mathbf{M}) denote the matrix whose (i,j)(i,j)-th entry is P[a,b](mij)\mathcal{P}_{[a,b]}(m_{ij}). For a constant C>0C>0, let the projection operator PC,2→∞\mathcal{P}_{C,2\to\infty} map an n×pn\times p matrix onto the constraint set {A∈Rn×p:∥A∥2→∞≤C}\{\mathbf{A}\in\mathbb{R}^{n\times p}:\|\mathbf{A}\|_{2\to\infty}\le C\}. For a matrix V=(v1,…,vn)T∈Rn×p\mathbf{V}=(\mathbf{v}_1,\ldots,\mathbf{v}_n)^{\mathsf{T}}\in\mathbb{R}^{n\times p}, PC,2→∞(V)=(v~1,…,v~n)T\mathcal{P}_{C,2\to\infty}(\mathbf{V})=(\widetilde{\mathbf{v}}_1,\ldots,\widetilde{\mathbf{v}}_n)^{\mathsf{T}}, where v~i=vi\widetilde{\mathbf{v}}_i=\mathbf{v}_i if ∥vi∥≤C\|\mathbf{v}_i\|\le C and v~i=(C/∥vi∥)vi\widetilde{\mathbf{v}}_i=(C/\|\mathbf{v}_i\|)\mathbf{v}_i otherwise. Additionally, let Jn=In−1n1nT/n\mathbf{J}_n=\mathbf{I}_n-\mathbf{1}_n\mathbf{1}_n^{\mathsf{T}}/n denote the centering matrix and we sometimes omit the subscript when the dimension is clear from matrix multiplication. Let OK\mathbb{O}_K denote the set of all KK-by-KK orthogonal matrices. For a square matrix M\mathbf{M}, let Δ(M)\Delta(\mathbf{M}) denote the matrix whose off-diagonal entries are the same as those of M\mathbf{M} while the diagonal entries are set to 00. Let Ba(K)(C)B_{\mathbf{a}}^{(K)}(C) denote the ball in RK\mathbb{R}^{K} centered at a∈RK\mathbf{a}\in\mathbb{R}^{K} with radius CC. The ψ1\psi_1-Orlicz norm of a random variable XX, where ψ1(x)=exp⁡(x)−1\psi_1(x)=\exp(x)-1, is defined as ∥X∥ψ1:=inf⁡{t>0:E[exp⁡(∣X∣/t)]≤2}\|X\|_{\psi_1}:=\inf\{t>0:\mathbb{E}[\exp(|X|/t)]\le2\}. Note that XX is sub-exponential if ∥X∥ψ1<∞\|X\|_{\psi_1}<\infty. Throughout the paper, c>0c>0 denotes a generic constant independent of (N,J)(N,J), whose value may change from line to line.

The remainder of the paper is organized as follows. Section 2 introduces the generalized distance-based latent space models and their connections to existing low-dimensional representation methods. Section 3 presents the proposed estimation procedure, including refined spectral denoising and configuration recovery. Section 4 establishes the corresponding theoretical guarantees for distance-signal estimation and configuration recovery. Section 5 presents simulation studies, and Section 6 applies the proposed method to a citation network of statisticians and European Parliament roll-call data. Section 7 concludes with a discussion, and the appendices contain proofs and additional numerical results.

Generalized Distance-Based Latent Space Models

To describe the unipartite and bipartite settings within a common notation, let P1\mathcal{P}_1 and P2\mathcal{P}_2 denote the object sets associated with the row and column indices of the observed relational matrix, respectively, and let P=P1∪P2\mathcal{P} = \mathcal{P}_1 \cup\mathcal{P}_2 denote the global object set. Define N=∣P1∣N = \lvert\mathcal{P}_1\rvert and J=∣P2∣J = \lvert\mathcal{P}_2\rvert. In the unipartite setting, P1=P2=[N]\mathcal{P}_1 = \mathcal{P}_2 = [N], whereas in the bipartite setting, P1=[N]\mathcal{P}_1 = [N] and P2=N+[J]={N+1,…,N+J}\mathcal{P}_2 = N + [J] = \{N + 1, \ldots, N + J\}.

Each object u∈Pu \in\mathcal{P} is associated with an embedding vector θu∈RK\boldsymbol{\theta}_u \in\mathbb{R}^K. We collect all embedding vectors into a global configuration Θ=(θu)u∈P\boldsymbol{\Theta} = (\boldsymbol{\theta}_u)_{u \in\mathcal{P}}. For g∈{1,2}g \in\{1, 2\}, let ΘPg=(θu)u∈Pg\boldsymbol{\Theta}_{\mathcal{P}_g} = (\boldsymbol{\theta}_u)_{u \in\mathcal{P}_g} denote the configuration consisting of the objects in Pg\mathcal{P}_g. We use uu and vv to denote global object indices in P\mathcal{P}, whereas ii and jj denote the local row and column indices of the observed relational matrix Y=(yij)\mathbf{Y} = (y_{ij}). The correspondence between the local indices (i,j)(i,j) and the associated global objects depends on the observation structure and will be specified separately for the unipartite and bipartite settings.

For each observed entry yijy_{ij}, let dijd_{ij} denote the squared Euclidean distance between the two corresponding global objects, and define mijm_{ij} as

mij=τ−dijwithdij={∥θi−θj∥2,unipartite,∥θi−θN+j∥2,bipartite,m_{ij}=\tau-d_{ij}\quad\text{with}\quad d_{ij}= \begin{cases} \lVert\theta_i-\theta_j\rVert^2, & \text{unipartite},\\ \lVert\theta_i-\theta_{N+j}\rVert^2, & \text{bipartite}, \end{cases}

where τ∈R\tau\in\mathbb{R} is an unknown drift parameter. We treat the embedding vectors as fixed unknown parameters and assume that the observations satisfy an independence structure appropriate to the underlying relational setting, as specified in Sections 2.1 and 2.2. We assume

E(yij)=f(mij)=f(τ−dij),(1)\mathbb{E}(y_{ij})=f(m_{ij})=f(\tau-d_{ij}), \tag*{(1)}

where f:R→Rf:\mathbb{R}\to\mathbb{R} is a prespecified strictly increasing link function. We refer to M=(mij)\mathbf{M}=(m_{ij}) as the signal matrix. Under (1), a smaller latent distance corresponds to a larger expected relation. The parameter τ\tau controls the overall level of the observed responses. If the observed entries represent dissimilarities rather than similarities, one may instead model E(yij)=f(−mij)\mathbb{E}(y_{ij})=f(-m_{ij}). For simplicity, we focus on the formulation in (1) throughout the paper.

The mean formulation in (1) accommodates different response types through the choice of ff. For continuous responses, one may take ff to be the identity function and regard yijy_{ij} as a noisy observation of mijm_{ij}. For binary responses, one may take ff to be a logistic or probit function, so that the probability of observing a positive relation decreases with the latent distance. Ordinal responses can be handled analogously through an ordinal response model. For count data, one may take f(x)=exp⁡(x)f(x)=\exp(x), leading to a Poisson mean model with intensity exp⁡(mij)\exp(m_{ij}).

References

  1. [1]Abbe, E., Fan, J., Wang, K., and Zhong, Y. (2020). Entrywise eigenvector analysis of random matrices with low expected rank. The Annals of Statistics, 48(3):1452–1474.DOI
  2. [2]Agarwal, A., Negahban, S., and Wainwright, M. J. (2012). Noisy matrix decomposition via convex relaxation: Optimal rates in high dimensions. The Annals of Statistics, 40(2):1171–1197.
  3. [3]Bai, J. (2003). Inferential theory for factor models of large dimensions. Econometrica, 71(1):135–171.DOI
  4. [4]Bandeira, A. S. and van Handel, R. (2016). Sharp nonasymptotic bounds on the norm of random matrices with independent entries. The Annals of Probability, 44(4):2479–2506.DOI
  5. [5]Bartholomew, D. J., Steele, F., Galbraith, J., and Moustaki, I. (2008). Analysis of multivariate social science data. CRC press.DOI
  6. [6]Belkin, M. and Niyogi, P. (2003). Laplacian eigenmaps for dimensionality reduction and data representation. Neural Computation, 15(6):1373–1396.DOI
  7. [7]Borg, I. and Groenen, P. J. F. (2005). Modern Multidimensional Scaling: Theory and Applications. Springer, New York, 2 edition.
  8. [8]Busing, F. M., Groenen, P. J., and Heiser, W. J. (2005). Avoiding degeneracy in multidimensional unfolding by penalizing on the coefficient of variation. Psychometrika, 70:71–98.DOI
  9. [9]Chatterjee, S. (2015). Matrix estimation by universal singular value thresholding. The Annals of Statistics, 43(1):177 – 214.DOI
  10. [12]Chen, Y., Li, X., and Zhang, S. (2020). Structured latent factor analysis for large-scale data: Identifiability, estimability, and their implications. Journal of the American Statistical Association, 115(532):1756–1770.arxiv.org/abs/1712.08966
  11. [13]Chen, Y., Ying, Z., and Zhang, H. (2021). Unfolding-model-based visualization: Theory, method and applications. Journal of Machine Learning Research, 22(11):1–51.arxiv.org/abs/2009.01551
  12. [14]de Bodt, C., Diaz-Papkovich, A., Bleher, M., Bunte, K., Coupette, C., Damrich, S., Fita Sanmartin, E., Hamprecht, F. A., Horvát, E.-A., Kohli, D., Krishnaswamy, S., Lee, J. A., Lelieveldt, B. P. F., McInnes, L., Nabney, I. T., Noichl, M., Polícar, P. G., Rieck, B., Wolf, G., Mishne, G., and Kobak, D. (2025). Low-dimensional embeddings of high-dimensional data.arxiv.org/abs/2508.15929
  13. [15]de Leeuw, J. and Mair, P. (2009). Multidimensional scaling using majorization: SMACOF in R. Journal of Statistical Software, 31(3):1–30.
  14. [16]Fan, J., Liao, Y., and Wang, W. (2016). Projected principal component analysis in factor models. The Annals of Statistics, 44(1):219–254.DOI
  15. [17]Gabriel, K. R. (1971). The biplot graphic display of matrices with application to principal component analysis. Biometrika, 58(3):453–467.DOI
  16. [18]Gabriel, K. R. (2002). Goodness of fit of biplots and correspondence analysis. Biometrika, 89(2):423–436.DOI
  17. [19]Gower, J. C. (1966). Some distance properties of latent root and vector methods used in multivariate analysis. Biometrika, 53(3/4):325–338.DOI
  18. [20]Greco, S. (2025). Two-dimensional political competition and europarties as cleavage mediators in the 2024 european parliament elections. JCMS: Journal of Common Market Studies, 63(6):1507–1525.DOI
  19. [22]Hinton, G. E. and Salakhutdinov, R. R. (2006). Reducing the dimensionality of data with neural networks. Science, 313(5786):504–507.pubmed.ncbi.nlm.nih.gov/16873662
  20. [23]Hix, S., Noury, A., and Roland, G. (2006). Dimensions of politics in the european parliament. American Journal of Political Science, 50(2):494–520.DOI
  21. [24]Hoff, P. D., Raftery, A. E., and Hancock, M. S. (2002). Latent space approaches to social network analysis. Journal of the American Statistical Association, 97(460):1090–1098.
  22. [25]Kidziński, Ł., Hui, F. K. C., Warton, D. I., and Hastie, T. J. (2022). Generalized matrix factorization: Efficient algorithms for fitting generalized linear latent variable models to large data arrays. Journal of Machine Learning Research, 23(291):1–29.
  23. [26]Kruskal, J. B. (1964). Multidimensional scaling by optimizing goodness of fit to a nonmetric hypothesis. Psychometrika, 29:1–27.DOI
  24. [27]Latala, R. (2005). Some estimates of norms of random matrices. Proceedings of the American Mathematical Society, 133(5):1273–1282.DOI
  25. [28]Lee, S. and Gu, Y. (2026). Deep discrete encoders: Identifiable deep generative models for rich data with discrete latent layers. Journal of the American Statistical Association, 121(553):194–208.arxiv.org/abs/2501.01414
  26. [31]Little, A., Xie, Y., and Sun, Q. (2023). An analysis of classical multidimensional scaling with applications to clustering. Information and Inference: A Journal of the IMA, 12(1):72–112.DOI
  27. [32]Lyu, Z. and Gu, Y. (2026). Spectral clustering with likelihood refinement for high-dimensional latent class recovery. Psychometrika, 91(2):695–723.DOI
  28. [34]McInnes, L., Healy, J., and Melville, J. (2018). UMAP: Uniform manifold approximation and projection for dimension reduction.arxiv.org/abs/1802.03426
  29. [35]Schönemann, P. H. (1970). On metric multidimensional unfolding. Psychometrika, 35(3):349–366.
  30. [36]Skrondal, A. and Rabe-Hesketh, S. (2004). Generalized latent variable modeling: Multilevel, longitudinal, and structural equation models. Chapman and Hall/CRC.
  31. [37]Tenenbaum, J. B., de Silva, V., and Langford, J. C. (2000). A global geometric framework for nonlinear dimensionality reduction. Science, 290(5500):2319–2323.pubmed.ncbi.nlm.nih.gov/11125149
  32. [38]Torgerson, W. S. (1952). Multidimensional scaling: I. theory and method. Psychometrika, 17(4):401–419.DOI
  33. [39]van der Maaten, L. and Hinton, G. (2008). Visualizing data using t-SNE. Journal of Machine Learning Research, 9:2579–2605.
  34. [40]Vershynin, R. (2018). High-Dimensional Probability: An Introduction with Applications in Data Science. Number 47 in Cambridge Series in Statistical and Probabilistic Mathematics. Cambridge University Press.
  35. [41]Wang, F. (2022). Maximum likelihood estimation and inference for high dimensional generalized factor models with application to factor-augmented regressions. Journal of Econometrics, 229(1):180–200.
  36. [42]Xia, D., Yuan, M., and Zhang, C.-H. (2021). Statistically optimal and computationally efficient low rank tensor completion from noisy entries. The Annals of Statistics, 49(1):76–99.arxiv.org/abs/1711.04934
  37. [43]Zhang, H., Chen, Y., and Li, X. (2020). A note on exploratory item factor analysis by singular value decomposition. Psychometrika, 85(2):358–372.DOI
  38. [44]Zhang, L., Wahba, G., and Yuan, M. (2016). Distance shrinkage and euclidean embedding via regularized kernel estimation. Journal of the Royal Statistical Society: Series B (Statistical Methodology), 78(4):849–867.arxiv.org/abs/1409.5009

Paper details

vlnc:W_pqewvh4b
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Mathematics

Not recorded for this paper

View statement
Looking for related papers…
Contents