betaFeedback

WorksComputer Science

Metric Testing: A Characterization of One-Sided Testability

Abstract

We give a unified framework for one-sided testability (constant-query testing) of metric families. For a broad class of families defined by fixed local predicates, we prove that testability is equivalent to two structural properties: packing amplification, which turns many edge-disjoint local violations into a positive density of violations, and completion stability, which allows locally consistent partial matrices with few missing entries to be cheaply completed. This characterization yields a common proof strategy and simpler proofs of classical testability results for ultrametrics, tree pseudometrics, Euclidean pseudometrics in fixed dimension, and several others. Our framework also gives new quantitative bounds for metric testing: a nonadaptive one-sided upper bound for matrices with entries in the bounded range [1,C][1,C], and two-sided nonadaptive and adaptive lower bounds for general metrics.

Introduction

Given query access to a symmetric matrix D∈Rn×nD \in\mathbb{R}^{n \times n} of nonnegative entries with zero diagonal and a prescribed metric family P={Pn}\mathcal{P} = \{\mathcal{P}_n\}, metric property testing asks us to distinguish whether D∈PnD \in\mathcal{P}_n or DD is ϵ\epsilon-far from Pn\mathcal{P}_n. Here, ϵ\epsilon-far means that more than ϵn2/2\epsilon n^2/2 unordered off-diagonal pairs must be changed to obtain a member of Pn\mathcal{P}_n. The goal is to test the P\mathcal{P}-property using as few queries to the entries of DD as possible. Classic examples include testing whether DD satisfies the triangle inequality, is an ultrametric, or can be realized by shortest-path distances in a tree or a Euclidean space of fixed dimension [31, 27, 30, 9]. These questions have found numerous applications in metric recognition, completion, and repair [29, 32, 12, 8, 7, 21, 20, 19, 13, 2].

A one-sided ϵ\epsilon-tester is a randomized algorithm that accepts every member in Pn\mathcal{P}_n with probability 1 and rejects every ϵ\epsilon-far input with probability at least 2/3. We call the family P\mathcal{P} one-sided testable if, for every ϵ>0\epsilon> 0, there is such a tester whose query complexity is bounded by q(ϵ)q(\epsilon) independently of nn.

Parnas and Ron [31] initiated the study of metric property testing and proved one-sided testability for several families, including ultrametrics and non-injective tree and Euclidean metrics. Krauthgamer and Sasson [27] improved the testing bounds for finite dimensional Euclidean realizability. More recently, Bao, Kannan, and Waingarten [9] obtained improved bounds for ultrametric testing, studied tree metric testing, and gave a sublinear-query tester for general metrics. They also proved lower bounds for general metric testing, ruling out nonadaptive one-sided testers whose query complexity is polynomial in 1/ϵ1/\epsilon and independent of nn. Together, these results raise a structural question:

What structural properties make a metric family one-sided testable?

In fact, one-sided testability already forces locality: ϵ\epsilon-far inputs must reveal small certificates of non-membership under random sampling. Indeed, for families invariant under relabelling, an adaptation of the canonicalization argument of Goldreich and Trevisan [25] simulates a q(ϵ)q(\epsilon)-query tester using the induced matrix on a uniformly sampled set of at most 2q(ϵ)2q(\epsilon) points. Perfect completeness forces the simulation to accept every sampled configuration that occurs in a valid input. Thus, on an ϵ\epsilon-far input, the sample must contain, with probability at least 2/3, a configuration incompatible with every valid completion on nn points. Small, detectable obstructions are therefore a necessary consequence of one-sided testability. This perspective also underlies Alon and Shapira’s characterization of oblivious one-sided graph testing through semi-hereditary properties [5].

It is therefore natural to focus on metric families characterizable by a local predicate: a matrix belongs to the family exactly when every kk-point restriction satisfies the predicate, where kk depends only on the family. This structure is shared by many families: general metrics and ultrametrics are determined by 3-point conditions, tree realizability by conditions on at most 4 points, and Euclidean realizability in fixed dimension by conditions on a bounded number of points. For these families, the challenge is to understand when being far from membership makes local violations detectable by sampling.

Our main contribution is a unified structural characterization of one-sided testability. Our approach connects two tasks, repairing an input and detecting its violations, through two complementary properties:

  • Packing amplification: many independent violations force abundant violations.

  • Completion stability: few holes in locally consistent data require few edits.

To see how these properties work together, delete the pairs covered by a maximal collection of edge-disjoint local violations. Maximality ensures that the remaining partial matrix is locally consistent. If the collection is small, only a few holes are created, and completion stability yields a valid matrix close to the original input. Thus, an input far from the family must contain many edge-disjoint violations, which packing amplification turns into sufficiently many violations for random sampling to detect. We show that these two properties together characterize one-sided testability. Separating these tasks gives a common proof strategy across metric families, yields simplified proofs of several classical testability results, and identifies distinct mechanisms behind the failure of testability.

Our results

Fix a local predicate Φ\Phi on kk points characterizing a metric family P\mathcal{P}. A kk-point restriction that fails Φ\Phi is called a local violation. We introduce two properties, with all bounds uniform in n≥kn \ge k.

Packing amplification. For every α>0\alpha> 0, there exists δ>0\delta> 0 such that any nn-point matrix containing at least αn2\alpha n^2 edge-disjoint local violations contains at least δ(k)(n)\delta^{(n)}_{(k)} local violations in total. Here, edge-disjointness means that no two violations share an unordered pair of indices.

Completion stability. A partial matrix contains unspecified pairs, called holes, and is locally consistent if every fully specified kk-point restriction satisfies Φ\Phi. For every η>0\eta> 0, there exists θ>0\theta> 0 such that any locally consistent partial matrix containing at most θn2\theta n^2 holes can be completed to a member of Pn\mathcal{P}_n after changing at most ηn2\eta n^2 specified pairs (filling the holes incurs no cost).

Theorem 1.1 (Informal). For metric families characterized by a local predicate, one-sided testability is equivalent to packing amplification together with completion stability.

Formal definitions and the precise theorem statement are given in § 3.

Unified and simplified analysis for one-sided testability. We then apply this characterization to several metric families (see Table 1). Our framework recovers the qualitative one-sided testability results of [31, 27, 9] for ultrametrics, strict ultrametrics, tree pseudometrics, and Euclidean pseudometrics in fixed dimension, with unified and simpler proofs via the packing amplification and the completion stability properties. For ultrametrics, these properties follow from elementary counting and partitioning arguments. For tree pseudometrics, both properties reduce to their ultrametric counterparts. For Euclidean pseudometrics of fixed dimension, both arguments use the rigidity of distances to a small affine anchor set. Therefore, the same testing argument applies across families, and parts of the proofs can be reused. The same framework identifies distinct reasons for non-testability. General metrics and metrics with bounded aspect ratio (entries in [1,C][1,C] for fixed C>2C > 2) fail packing amplification. Strict tree and Euclidean metrics retain packing amplification but fail completion stability. Also, as metrics with bounded aspect ratio satisfy completion stability, these examples establish the independence of the two properties.

FamilyPacking amplificationCompletion stabilityPrevious work
UltrametricsYes (Lemma 4.1)Yes (Lemma 5.2)[31]
Strict ultrametricsYes (Corollary 4.9)Yes (Corollary 5.3)[9]
Tree pseudometricsYes (Lemma 4.4)Yes (Lemma 5.5)[31]
Euclidean pseudometrics in Rd\mathbb{R}^dYes (Lemma 4.6)Yes (Lemma 5.7)[31, 27]
Strict tree metricsYes (Corollary 4.9)No (Theorem 5.20)[31]
Strict Euclidean metrics in Rd\mathbb{R}^dYes (Corollary 4.9)No (Theorem 5.20)[31]
Bounded aspect ratio metricsNo (Theorem 4.11)Yes (Theorem 5.13)—
General metricsNo (Theorem 4.11)—[9]

Table 1. Packing amplification and completion stability for the metric families studied in this paper. The dimension d≥1d \ge1 is a fixed constant. Strictness requires positive distances between distinct labels. For metrics with bounded aspect ratio, off-diagonal entries lie in [1,C][1,C] for a fixed C>2C > 2.

Quantitative testing bounds. Beyond giving a qualitative classification, our structural framework also gives new quantitative upper and lower bounds. Because our characterization reduces one-sided testability to two combinatorial properties, bounding completion stability and packing amplification yields explicit query upper bounds. For ultrametrics, strict ultrametrics, and tree pseudometrics, combining linear completion stability with quadratic packing amplification yields one-sided O(ϵ−2)O(\epsilon^{-2})-query testers (Corollary 6.1), removing the logarithmic factors from the O~(ϵ−2)\widetilde{O}(\epsilon^{-2}) bounds of [9]. We also provide novel bounds for general nn-point metrics and metrics with entries in [1,C][1,C]. To state the full tradeoff across all proximity regimes, define the rate function

F(n,ϵ)={n2,0<ϵ≤n−2,n/ϵ,n−2<ϵ≤n−1,n/ϵ,n−1<ϵ≤n−1/2,(n/ϵ)2/3,n−1/2<ϵ≤1.F(n,\epsilon) = \begin{cases} n^2, & 0 < \epsilon\le n^{-2},\\ n/\sqrt{\epsilon}, & n^{-2} < \epsilon\le n^{-1},\\ \sqrt{n}/\epsilon, & n^{-1} < \epsilon\le n^{-1/2},\\ (n/\epsilon)^{2/3}, & n^{-1/2} < \epsilon\le1. \end{cases}

Theorem 1.2. Fix C>2C > 2 and ϵ0=1/80\epsilon_0 = 1/80. For all sufficiently large nn and 0<ϵ≤ϵ00 < \epsilon\le\epsilon_0:

  1. Sharp nonadaptive complexity on [1,C][1,C] (Theorem 6.8): For inputs with entries in [1,C][1,C], both one-sided and two-sided nonadaptive ϵ\epsilon-testing of general metrics have query complexity ΘC(F(n,ϵ))\Theta_C(F(n,\epsilon)).

  1. Adaptive lower bound on [1,C][1,C] (Theorem 6.11): Any adaptive two-sided ϵ\epsilon-tester for general metrics on [1,C][1,C] requires ΩC(n/log⁡n)\Omega_C(\sqrt{n/\log n}) queries.

  1. Transfer to general and δ\delta-hyperbolic metrics: On arbitrary nonnegative inputs, nonadaptive and adaptive two-sided ϵ\epsilon-testing require Ω(F(n,ϵ))\Omega(F(n,\epsilon)) and Ω(n/log⁡n)\Omega(\sqrt{n/\log n}) queries, respectively, both for general metrics and pseudometrics (Theorem 6.12) and for δ\delta-hyperbolic metrics with any fixed δ>0\delta> 0 (Remark 4.12).

Previously, Bao, Kannan, and Waingarten [9] gave a nonadaptive one-sided query lower bound of Ω(n2/3−o(1))\Omega(n^{2/3-o(1)}) using Behrend graphs that applied only at vanishing proximity ϵ=n−o(1)\epsilon= n^{-o(1)}. Because our packing obstruction avoids progression-free sets, Theorem 1.2 eliminates the no(1)n^{o(1)} loss and establishes hardness at fixed constant ϵ\epsilon even against two-sided and adaptive testers.

Related work, and comparison with our results

Parnas and Ron [31] first introduced metric property testing, giving one-sided testers with O(ϵ−6)O(\epsilon^{-6}) queries for tree pseudometrics and ultrametrics, and O(d2log⁡2(d+1)/ϵ2)O(d^2\log^2(d+1)/\epsilon^2) queries for Euclidean pseudometrics in Rd\mathbb{R}^d. Krauthgamer and Sasson [27] improved the latter bound to O(d2/ϵ2)O(d^2/\epsilon^2) using O(d/ϵ)O(d/\epsilon) sampled points. For strict tree metrics and injective Euclidean realizations, Parnas and Ron proved Ω(n)\Omega(\sqrt{n}) query lower bounds even with two-sided error. Our proofs unify these qualitative results; our Euclidean bound has weaker ϵ\epsilon-dependence than [27].

Bao, Kannan, and Waingarten [9] established a nonadaptive one-sided O(n2/3/ϵ4/3)O(n^{2/3}/\epsilon^{4/3})-query tester for general metrics and improved the ultrametric testing bound to O~(ϵ−2)\widetilde{O}(\epsilon^{-2}). They also studied tree metric testing. Their Behrend graph lower bound for general metrics applies at ϵ\epsilon vanishing with nn. Our obstruction holds at fixed ϵ\epsilon, even for entries in [1,C][1,C] with any fixed C>2C > 2. Onak [30] studied tree metrics, ultrametrics, and line embeddings in a model measuring distance by point modifications.

For graphs, Goldreich and Trevisan [25] canonicalized constant-query testers using uniformly sampled induced subgraphs. Alon and Shapira [5] established one-sided testability of every hereditary graph property and characterized oblivious one-sided testability. Austin and Tao [6] developed testing and repair results for hereditary graph and hypergraph properties; Alon, Ben-Eliezer, and Fischer [1] established testability for hereditary ordered graphs and finite-alphabet matrices. We decompose removability for real-valued matrices into two structural conditions; completion stability concerns existence of close completions and differs from the local repairability in [6].

Metric repair and the metric violation distance problem also aim at minimizing the number of changed entries [21, 20] of an input matrix. Fan, Gilbert, Raichel, Sonthalia, and Van Buskirk [19] studied incomplete graphs using nonmetric cycles. Cohen-Addad, Fan, Lee, and de Mesmay [13] obtained an O(log⁡n)O(\log n)-approximation for metric violation distance and a constant-factor approximation for ultrametric violation distance. Bandyapadhyay, Lochet, Lokshtanov, Saurabh, Sonar, Surianarayanan, and Xue [10] gave a single-exponential fixed-parameter algorithm and a linear kernel for metric repair, and proved fixed-parameter tractability of tree metric repair. Our bounded range completion theorem controls the edits needed to fill sparse holes, including those forced by longer nonmetric cycles. Metric nearness instead minimizes a norm of the numerical perturbation [7].

Organization of the paper

§ 2 introduces the testing model, local predicates, and the metric families considered in the paper. § 3 proves the structural characterization of one-sided testability through removal, packing amplification, and completion stability. § 4 establishes packing amplification for the geometric families and constructs a packing obstruction for general metrics. § 5 proves the corresponding completion results and identifies the obstruction caused by injective realizations. Finally, § 7 derives quantitative testing bounds, including sharp nonadaptive bounds for metrics with bounded aspect ratio and an adaptive lower bound.

Preliminaries

Fix an integer k≥2k \ge2 and a nonempty value set Λ⊆R≥0\Lambda\subseteq\mathbb{R}_{\ge0}. For n≥1n \ge1, let Dn\mathcal{D}_n be the set of symmetric matrices with off-diagonal entries in Λ\Lambda and diagonal entries fixed to 00.

For A,B∈DnA,B \in\mathcal{D}_n, define the normalized Hamming distance between A,BA,B as

dist⁡n(A,B)=2n2∣{e∈([n]2):Ae≠Be}∣.\operatorname{dist}_n(A,B) = \frac{2}{n^2}\left|\left\{e \in\binom{[n]}{2}: A_e \ne B_e\right\}\right|.

For a nonempty set Q⊆Dn\mathcal{Q} \subseteq\mathcal{D}_n, define dist⁡n(A,Q)=min⁡B∈Qdist⁡n(A,B)\operatorname{dist}_n(A,\mathcal{Q}) = \min_{B \in\mathcal{Q}} \operatorname{dist}_n(A,B).

For probability distributions μ,ν\mu,\nu, denote by TV⁡(μ,ν)=sup⁡E∣μ(E)−ν(E)∣\operatorname{TV}(\mu,\nu) = \sup_E |\mu(E)-\nu(E)| the total variation distance between μ,ν\mu,\nu, where the supremum is over measurable events.

Local Predicates and Testability

Let Φ⊆Λ([k]2)\Phi\subseteq\Lambda^{\binom{[k]}{2}} be a permutation-invariant kk-point predicate. For S∈([n]k)S \in\binom{[n]}{k}, membership of D[S]D[S] in Φ\Phi is interpreted after pulling D[S]D[S] back along any bijection [k]→S[k] \to S.

Define

Pn={D∈Dn:D[S] satisfies Φ for every S∈([n]k)}.\mathcal{P}_n=\left\{D\in\mathcal{D}_n:D[S]\text{ satisfies }\Phi\text{ for every }S\in\binom{[n]}{k}\right\}.

For n<kn<k, Pn=Dn\mathcal{P}_n=\mathcal{D}_n. Unless stated otherwise, we assume Pn≠∅\mathcal{P}_n\ne\varnothing for every n≥kn\ge k.

For D∈DnD\in\mathcal{D}_n with n≥kn\ge k, define

V(D)={S∈([n]k):D[S] violates Φ},andviol⁡(D)=∣V(D)∣(nk).\mathcal{V}(D)=\left\{S\in\binom{[n]}{k}:D[S]\text{ violates }\Phi\right\},\qquad\text{and}\qquad\operatorname{viol}(D)=\frac{|\mathcal{V}(D)|}{\binom{n}{k}}.

Definition 2.1 (One-sided tester). A one-sided qq-query ϵ\epsilon-tester for P={Pn}\mathcal{P}=\{\mathcal{P}_n\} is an information-theoretic randomized adaptive decision tree which queries at most qq off-diagonal entries and satisfies:

(1) every D∈PnD\in\mathcal{P}_n is accepted with probability 11;

(2) every DD with dist⁡n(D,Pn)>ϵ\operatorname{dist}_n(D,\mathcal{P}_n)>\epsilon is rejected with probability at least 2/32/3.

References

  1. [1]Noga Alon, Omri Ben-Eliezer, and Eldar Fischer. “Testing Hereditary Properties of Ordered Graphs and Matrices”. In: Proceedings of the IEEE Symposium on Foundations of Computer Science (FOCS). 2017, pp. 848–858. arXiv: 1704.02367.arxiv.org/abs/1704.02367
  2. [2]A. Aranda, D. Bradley-Williams, E. K. Hng, J. Hubička, M. Karamanlis, M. Kompatscher, M. Konečný, and M. Pawliuk. “Completing graphs to metric spaces”. In: Contributions to Discrete Mathematics 16.2 (2021), pp. 71–89.
  3. [3]Noga Alon, Tali Kaufman, Michael Krivelevich, and Dana Ron. “Testing Triangle-Freeness in General Graphs”. In: SIAM Journal on Discrete Mathematics 22.2 (2008), pp. 786–819. DOI: 10.1137/07067917X.DOI
  4. [4]Noga Alon. “On the number of subgraphs of prescribed type of graphs with a given number of edges”. In: Israel Journal of Mathematics 38.1–2 (1981), pp. 116–130. DOI: 10.1007/BF02761855.
  5. [5]Noga Alon and Asaf Shapira. “A characterization of the (natural) graph properties testable with one-sided error”. In: SIAM Journal on Computing 37.6 (2008), pp. 1703–1727.
  6. [6]Tim Austin and Terence Tao. “Testability and repair of hereditary hypergraph properties”. In: Random Structures & Algorithms 36.4 (2010), pp. 373–463.
  7. [7]Justin Brickell, Inderjit S. Dhillon, Suvrit Sra, and Joel A. Tropp. “The metric nearness problem”. In: SIAM Journal on Matrix Analysis and Applications 30.1 (2008), pp. 375–396.DOI
  8. [8]Mihály Bakonyi and Charles R. Johnson. “The Euclidean distance matrix completion problem”. In: SIAM Journal on Matrix Analysis and Applications 16.2 (1995), pp. 646–654.
  9. [9]Yiqiao Bao, Sampath Kannan, and Erik Waingarten. “Nearly tight bounds on testing of metric properties”. In: Proceedings of the ACM–SIAM Symposium on Discrete Algorithms (SODA). 2025, pp. 403–445. arXiv: 2411.08989.
  10. [10]Sayan Bandyapadhyay, William Lochet, Daniel Lokshtanov, Saket Saurabh, Chinmay Sonar, Vaishali Surianarayanan, and Jie Xue. How to Fix a Broken Metric: A Linear Kernel, Tight Bounds, and Tree Metrics. 2026. arXiv: 2610.07690 [cs.DS].
  11. [11]L. M. Blumenthal. Theory and Applications of Distance Geometry. Oxford: Clarendon Press, 1953.
  12. [12]Peter Buneman. “A note on the metric properties of trees”. In: Journal of Combinatorial Theory, Series B 17.1 (1974), pp. 48–50.
  13. [13]Vincent Cohen-Addad, Chenglin Fan, Euiwoong Lee, and Arnaud de Mesmay. “Fitting metrics and ultrametrics with minimum disagreements”. In: Proceedings of the IEEE Symposium on Foundations of Computer Science (FOCS). 2022, pp. 301–311.
  14. [14]Fan R. K. Chung, Ronald L. Graham, Peter Frankl, and James B. Shearer. “Some intersection theorems for ordered sets and graphs”. In: Journal of Combinatorial Theory, Series A 43.1 (1986), pp. 23–37.
  15. [15]I. Csiszár. “I-divergence geometry of probability distributions and minimization problems”. In: Annals of Probability 3.1 (1975), pp. 146–158.
  16. [16]M. M. Deza and M. Laurent. Geometry of Cuts and Metrics. Springer, 1997.
  17. [17]John N. Darroch, Steffen L. Lauritzen, and Terry P. Speed. “Markov fields and log-linear interaction models for contingency tables”. In: Annals of Statistics 8.3 (1980), pp. 522–539.
  18. [18]James S. Farris. “Estimating phylogenetic trees from distance matrices”. In: The American Naturalist 106.951 (1972), pp. 645–668. DOI: 10.1086/282802.
  19. [19]Chenglin Fan, Anna C. Gilbert, Benjamin Raichel, Rishi Sonthalia, and Gregory Van Buskirk. “Generalized metric repair on graphs”. In: Proceedings of the 17th Scandinavian Symposium and Workshops on Algorithm Theory (SWAT). Vol. 162. Leibniz International Proceedings in Informatics (LIPIcs). 2020, 25:1–25:16. arXiv: 1908.08411.arxiv.org/abs/1908.08411
  20. [20]Chenglin Fan, Benjamin Raichel, and Gregory Van Buskirk. “Metric violation distance: Hardness and approximation”. In: Algorithmica 84.5 (2022), pp. 1441–1465.
  21. [21]A. C. Gilbert and L. Jain. “If it ain’t broke, don’t fix it: Sparse metric repair”. In: Proceedings of the 55th Annual Allerton Conference on Communication, Control, and Computing. 2017, pp. 612–619.arxiv.org/abs/1710.10655
  22. [22]M. Grötschel, L. Lovász, and A. Schrijver. Geometric Algorithms and Combinatorial Optimization. Springer, 1988.
  23. [23]Oded Goldreich and Dana Ron. “On proximity-oblivious testing”. In: SIAM Journal on Computing 40.2 (2011), pp. 534–566.
  24. [24]Mikhael Gromov. “Hyperbolic groups”. In: Essays in Group Theory. Ed. by S. M. Gersten. Vol. 8. Mathematical Sciences Research Institute Publications. New York: Springer, 1987, pp. 75–263. DOI: 10.1007/978-1-4613-9586-7_3.DOI
  25. [25]Oded Goldreich and Luca Trevisan. “Three theorems regarding testing graph properties”. In: Random Structures & Algorithms 23.1 (2003), pp. 23–57.
  26. [26]S. Louis Hakimi and Stephen S. Yau. “Distance matrix of a graph and its realizability”. In: Quarterly of Applied Mathematics 22.4 (1965), pp. 305–317.DOI
  27. [27]R. Krauthgamer and O. Sasson. “Property testing of data dimensionality”. In: Proceedings of the ACM–SIAM Symposium on Discrete Algorithms. 2003, pp. 18–27.DOI
  28. [28]S. L. Lauritzen. Graphical Models. Oxford University Press, 1996.
  29. [29]Karl Menger. “New foundations of Euclidean geometry”. In: American Journal of Mathematics 53.4 (1931), pp. 721–745.DOI
  30. [30]Krzysztof Onak. “Testing Properties of Sets of Points in Metric Spaces”. In: Automata, Languages and Programming (ICALP). Vol. 5125. Lecture Notes in Computer Science. Springer, 2008, pp. 515–526. DOI: 10.1007/978-3-540-70575-8_42.DOI
  31. [31]M. Parnas and D. Ron. “Testing metric properties”. In: Information and Computation 187.2 (2003), pp. 155–195.
  32. [32]I. J. Schoenberg. “Remarks to Maurice Fréchet’s article “Sur la définition axiomatique d’une classe d’espace distanciés vectoriellement applicable sur l’espace de Hilbert””. In: Annals of Mathematics 36.3 (1935), pp. 724–732.
  33. [33]A. Sidiropoulos, D. Wang, and Y. Wang. Metric embeddings with relaxed guarantees. 2015. arXiv: 1508.03600.
  34. [34]Martin J. Wainwright and Michael I. Jordan. “Graphical models, exponential families, and variational inference”. In: Foundations and Trends in Machine Learning 1.1–2 (2008), pp. 1–305.
  35. [35]Gale Young and Alston S. Householder. “Discussion of a set of points in terms of their mutual distances”. In: Psychometrika 3.1 (1938), pp. 19–22. DOI: 10.1007/BF02287916.DOI

Paper details

vlnc:W_x3p2s5ws
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.

  • Computer Science

Not recorded for this paper

View statement
Looking for related papers…
Contents