Pith. sign in

REVIEW 5 major objections 4 minor 47 references

KNN and K-means in Gini Prametric Spaces

T0 review · 5 major / 4 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read Replacing metric distances with a Gini prametric—a value-rank hybrid—is claimed to make K-means and KNN robust to noise, with Gini K-means provably convergent when rank vectors stay constant.

desk verdict Gini distance is worth a look, but the convergence proof is wrong and the K-means win is an oracle result. read the letter →

arxiv 2501.18028 v3 pith:ALYDN4U3 submitted 2025-01-29 cs.LG

Cassandra Mussard , Arthur Charpentier , Stéphane Mussard This is my paper · ORCID
classification cs.LG MSC 62H3068T10
keywords GiniprametricK-meansK-nearestneighborsrank-baseddistancerobustclusteringspacenoiserobustnessgeneralizedindex
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

The paper proposes replacing standard metric distances in K-means and K-nearest neighbors with a Gini prametric, a symmetric non-distance that multiplies coordinate differences by differences of powered ranks. It claims this rank-plus-value measure is robust to noise and outliers, and that the resulting Gini K-means converges whenever rank vectors stay constant for any $\nu \neq 1$. In experiments on sixteen public benchmark datasets, the generalized Gini KNN ranks at or near the top in precision and recall, matching or exceeding the Hassanat distance, which the paper treats as the strongest noise-robust baseline. If these claims hold, rank-based prametrics would be a practical drop-in replacement for metric distances in clustering and classification, giving provable convergence and noise resistance without the triangle inequality.

What carries the argument

The load-bearing object is the generalized Gini prametric $d_{G,\nu}$, a value-rank hybrid in which each coordinate gap $(x_{i,j}-x_{k,j})$ is weighted by the gap between powered descending ranks $R_{X_j}(\cdot)^{\nu-1}$. The hyperparameter $\nu$ tunes sensitivity to the tail of the feature distribution: $\nu=2$ treats upper and lower ranks symmetrically, while larger $\nu$ emphasizes higher ranks and smaller $\nu$ emphasizes lower ranks. This object carries the argument because it injects robustness through ranks while still reacting to actual value differences, and its invariance properties (linear and rank invariance) are what the convergence and robustness arguments rely on.

What would settle it

Evaluate the within-cluster objective $f(z)=\sum_i d_{G,\nu}(x_i,z)^2$ for a small one-dimensional cluster such as $\{0,2,10\}$ with fixed ranks $(1,2,3)$, and compare $z=4$ (the arithmetic mean) with a weighted mean derived from the first-order condition. If the weighted mean gives a strictly smaller value of $f$, the centroid-update lemma behind Proposition 4 is false; conversely, if the arithmetic mean always wins, the convergence proof is on solid ground.

Watch

Extended reading notes

Core claim

The central claim is that the generalized Gini prametric $d_{G,\nu}(x_i,x_k) = -\sum_{j=1}^d (x_{i,j}-x_{k,j})(R_{X_j}(x_{i,j})^{\nu-1} - R_{X_j}(x_{k,j})^{\nu-1})$, where $R_{X_j}$ is the descending rank of feature $j$, is a valid dissimilarity for KNN and K-means even though it is not a metric. The paper proves in Proposition 4 that Gini K-means converges for all $\nu \neq 1$ when rank vectors stay constant, and reports that Gini KNN achieves the best average precision rank among twelve distances on clean data, with competitive performance under added Gaussian noise. The same construction is extended to agglomerative hierarchical clustering, where the generalized Gini prametric ranks first in precision on noisy data.

Load-bearing premise

The convergence theorem depends on the lemma that, with ranks held fixed, the arithmetic mean of a cluster is the unique minimizer of the sum of squared Gini distances to the cluster points; if the true minimizer is instead a weighted mean, the centroid update in the proof no longer matches the average and the convergence argument collapses.

Editorial extensions

If this is right

  • If Proposition 4 is correct, Gini K-means is guaranteed to stop for any $\nu\neq 1$ whenever the rank vectors of the data do not change during iteration, making rank-based robust clustering a provably convergent procedure.
  • Gini KNN inherits the standard nearest-neighbor asymptotic error bound (Proposition 3), so the rank-based distance adds robustness without sacrificing the classical consistency guarantee.
  • The hyperparameter $\nu$ gives a tunable robustness lever: practitioners can shift weight between lower and upper parts of each feature's distribution, with $\nu=2$ as the symmetric baseline.
  • Because the prametric is linear-invariant and rank-invariant, monotone transformations of the features leave the distances essentially unchanged, which the paper argues protects against outliers and measurement errors.

Reading between the lines

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

  • A natural repair if the centroid lemma fails is to replace the arithmetic mean by the weighted mean that the derivative condition actually selects, and to prove a Lloyd-style potential decrease for that update; this would salvage convergence for a modified Gini K-means.
  • The conditional-rank construction used for train-test splits suggests that streaming or online use of the Gini prametric would need an explicit rule for updating ranks as new points arrive, since distances between old points change under re-ranking.
  • One could test whether $\nu$ should be adapted to the noise level itself: the reported advantage over the Hassanat distance is clearest at 5% noise and shrinks at 10%, so a noise-dependent schedule for $\nu$ may extend the method's robustness.
  • The same value-rank weighting idea could be applied to other algorithms that only need pairwise dissimilarities, such as spectral clustering or distance-based outlier detection, though the paper only demonstrates K-means, KNN, and agglomerative clustering.
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

5 major / 4 minor

Summary. The paper defines a generalized Gini prametric dG,ν that combines value differences with rank differences between points, and uses it to build variants of K-nearest neighbors and K-means. The main claims are: (i) the Gini prametric is robust to noise and outliers; (ii) Gini K-means is provably convergent when rank vectors remain constant; and (iii) Gini KNN is competitive with Hassanat distance. The theoretical convergence claim is delegated to the Supplementary Materials, and the empirical evaluation is carried out on 16 UCI datasets with and without added noise, reporting precision and recall rankings for KNN, K-means, and agglomerative clustering.

Significance. If the convergence theorem were correct, the paper would offer a practical, provably convergent rank-based alternative for robust K-means and KNN, and the extensive UCI experiments would be a useful benchmark. The paper also ships supplementary tables and describes reproducible experimental machinery. However, the central convergence proof rests on an incorrect centroid-update lemma, and the headline K-means experiments select the hyperparameter ν using labeled test data. These two issues undermine the paper's main theoretical and empirical contributions as they currently stand.

major comments (5)
  1. [Supplementary Materials, Proof of Proposition 4] The proof's key lemma—that the arithmetic mean uniquely minimizes the within-cluster sum of squared Gini distances under fixed ranks—is false. The derivative equation shown in the proof, even if accepted, gives a weighted mean, not the unweighted arithmetic mean. Concretely, for the scalar cluster {0,1,4} with ν=2 and descending ranks (3,2,1), for any centroid z between 1 and 4 the objective is f(z)=z^2+(4-z)^2, whose minimizer is z=2, not the arithmetic mean 5/3. In addition, the differentiation step itself is algebraically incorrect: the squared distance is the square of a sum over j, so differentiating Σ_i [Σ_j (z_ij−z_j)(R_ij^{ν−1}−R(z_j)^{ν−1})]^2 does not produce the simple equation written in the proof. Since inequality (1) depends entirely on this centroid-minimizer lemma, Proposition 4 is not established.
  2. [Section 4.2, Proposition 4 and Algorithm 1] The convergence statement is conditional on 'rank vectors stay constant', but the algorithm does not enforce or verify this condition. The centroid is updated to the arithmetic mean of the cluster, and there is no reason that the ranks of the new centroid with respect to the cluster members remain unchanged. Thus, even if the centroid-minimizer lemma were correct, the proof would cover only a hypothetical fixed-rank iteration, not the algorithm actually implemented and evaluated in the experiments.
  3. [Section 5.2, K-means Experiments] The reported K-means advantage is obtained by selecting ν to maximize precision on the supervised UCI datasets, i.e., using the test labels. The text states: 'a grid search is performed over the hyper-parameter ν, selecting the value that maximizes the precision score.' This is an oracle selection that inflates the mean ranks in Tables 6 and 7. The paper's own later silhouette-based selection (Tables 8 and 9) drops Gini K-means to rank 4 in precision and rank 3 in recall, far below the headline ranks of 3.50 and 4.19. The empirical claim of superiority therefore depends on information that would not be available in a real unsupervised application.
  4. [Section 4.1, Proposition 3] The statement that the Cover-Hart asymptotic error bound applies unchanged to the generalized Gini prametric is not justified. Cover and Hart's result relies on the metric structure of the space, in particular the property that the nearest neighbor converges to the query point as the sample size grows. The Gini prametric does not satisfy the triangle inequality and can assign zero distance to distinct points with identical rank vectors, so the standard proof does not carry over. At minimum, a proof is needed; the sentence 'The proof is exactly the same as the one of Cover [9]' does not address these differences.
  5. [Definitions 2 and Algorithm 1] There is an internal inconsistency about the domain of ν. Definition 2 and the surrounding text require ν>1, and this condition is what guarantees non-negativity of dG,ν. However, Algorithm 1 searches over ν in [0.1;6] excluding 1, and Proposition 4 claims convergence for all ν≠1. For 0<ν<1, the exponent ν−1 is negative, so the generalized Gini prametric can be negative and is no longer a prametric in the sense used elsewhere in the paper. The grid search and the convergence statement should be restricted to the range where the distance is well defined.
minor comments (4)
  1. [Section 3.3] The text says 'it can be referred to as a Gini pramatric space'; the intended term is 'prametric space'.
  2. [References [9] and [10]] Reference [9] appears to be a duplicate of [10] and lists the author as 'H. Cover'; the correct citation is T. M. Cover and P. E. Hart, 'Nearest neighbor pattern classification', IEEE Transactions on Information Theory, 1967.
  3. [Algorithm 1] The line 'RX = conditional ranks of X_tr[-fold]' is unclear: the notation X_tr[-fold] is not defined, and the procedure for computing conditional ranks when a fold is held out should be specified precisely.
  4. [Section 4.1, Proposition 3] The phrase 'The KNN algorithm converges' is not defined; KNN is a classification rule without an iterative update, so the statement should be replaced with a precise consistency or asymptotic-error claim.

Circularity Check

2 steps flagged · score 6.0 of 10

K-means empirical advantage is partly an artifact of ν tuned to maximize the reported precision, and the convergence theorem is deferred to a self-cited supplementary proof.

  1. self citation load bearing [Section 4.2, Proposition 4 (proof pointer) and Reference [27]]
    "Proposition 4. (Convergence) The Gini K-means algorithm based on the generalized Gini prametric dG,ν is convergent whenever rank vectors stay constant, for all ν ̸= 1. Proof. See the Supplementary Materials [27]."

    The paper's headline theoretical claim—convergence of Gini K-means—is justified solely by a pointer to 'Supplementary Materials [27]', where [27] is the present arXiv paper by the same three authors. No independent theorem or external proof is supplied. The referenced supplementary proof then assumes the centroid-minimizer lemma, so the convergence result rests on the authors' own self-referential derivation rather than on an independently verifiable mathematical fact. This is a self-citation chain rather than a reduction by definition, but it is load-bearing.

  2. fitted input called prediction [Section 5.2, K-means Experiments (paragraph before Table 6)]
    "For Gini K-means, a grid search is performed over the hyper-parameter ν, selecting the value that maximizes the precision score. Indeed, since the UCI datasets are supervised, computing the silhouette criterion is not required."

    The reported top performance of Gini K-means (Table 6, mean rank 3.50) is obtained by choosing ν to maximize the same precision metric that is then reported as the algorithm's predictive performance. The precision values are therefore not independent predictions: ν is fitted to the evaluation labels, so the reported precision is by construction the maximum over the grid. The paper's own Table 8, where ν is selected by the unsupervised silhouette criterion, drops Gini K-means to rank 4, confirming that the headline advantage is an artifact of using the labeled data to select ν.

full rationale

The Gini prametric itself is defined independently (Definition 1) in terms of value differences and rank differences, so the distance is not defined in terms of the algorithm's output; I find no self-definitional circularity in the metric. The KNN experiments tune k and ν on labeled training data with cross-validation, which is standard supervised model selection, and the comparison against Hassanat's distance on UCI benchmarks is an external benchmark, so the KNN claim has independent content. Two elements do raise the circularity score. First, the convergence of Gini K-means (Proposition 4), the paper's central theoretical novelty, is justified only by 'See the Supplementary Materials [27]', where [27] is the same three authors' own arXiv paper; the referenced proof then assumes the centroid-minimizer lemma, so the load-bearing theorem is supported by a self-citation chain rather than by an independent mathematical argument. Second, in the K-means experiments the hyper-parameter ν is selected by maximizing the precision score on the labeled UCI data and that same precision is then reported as the method's performance; the reported advantage is therefore partly a selection artifact. I do not count the incorrect centroid-minimizer step as circularity—it is a correctness flaw in the supplementary proof—and I note the paper honestly discloses that silhouette-based ν selection drops Gini K-means to rank 4, which limits but does not remove the fitted-input concern.

Assumptions & free parameters 2 free parameters · 3 assumptions · 0 invented entities

The central claim relies on two mathematically fragile axioms in the convergence proof and on a hyperparameter nu that is fit to the evaluation data. No new physical entities are proposed; the Gini prametric is a new mathematical function rather than an invented entity.

free parameters (2)
  • nu (generalized Gini rank exponent) = per dataset, e.g., 2.70, 3.52, 1.22
    Grid-searched per dataset to maximize precision or silhouette; headline K-means results select nu with known labels, so comparisons reflect fitting.
  • k (number of neighbors in KNN) = grid 1..11 per dataset
    Selected by maximizing F1 or precision; standard practice, but part of the reported advantage is tuning-dependent.
assumptions (3)
  • ad hoc to paper Arithmetic mean minimizes sum of squared Gini distances under fixed ranks
    Used in the Supplementary proof of Proposition 4; algebraically unsupported and false in general.
  • ad hoc to paper Cover's asymptotic error bound for KNN applies unchanged to any prametric without triangle inequality
    Proposition 3 says the proof is exactly the same as Cover [9] but does not verify that Cover's argument survives without the triangle inequality.
  • ad hoc to paper Rank function is differentiable with respect to small changes in centroid values
    The derivative step in the convergence proof assumes ranks remain constant under small variations of z; this is not a differentiable function at data points.

how reviews work

0 comments
Cite this review

Pith. "Pith review of KNN and K-means in Gini Prametric Spaces." pith.science (2026). https://pith.science/paper/ALYDN4U3

@misc{pith2026250118028,
  author       = {Pith},
  title        = {Pith review of: KNN and K-means in Gini Prametric Spaces},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ALYDN4U3}},
  note         = {Machine review of arXiv:2501.18028}
}
read the original abstract

This paper introduces enhancements to the K-means and K-nearest neighbors (KNN) algorithms based on the concept of Gini prametric spaces, instead of traditional metric spaces. Unlike standard distance metrics, Gini prametrics incorporate both value-based and rank-based measures, offering robustness to noise and outliers. The main contributions include: (1) a Gini prametric that captures rank information alongside value distances; (2) a Gini K-means algorithm that is provably convergent and resilient to noisy data; and (3) a Gini KNN method that performs competitively with state-of-the-art approaches like Hassanat's distance in noisy environments. Experimental evaluations on 16 UCI datasets demonstrate the superior performance and efficiency of the Gini-based algorithms in clustering and classification tasks. This work opens new directions for rank-based prametrics in machine learning and statistical analysis.

Figures

Figures reproduced from arXiv: 2501.18028 by the authors.

Figure 2
Figure 2. Example of agglomerative clustering: ν = 2 Finally, other clustering techniques relying on the Euclidean dis￾tance, close to K-means, may be adapted to the generalized Gini prametric, such as the agglomerative clustering (see subsection 5.3). An illustration is depicted in [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 1
Figure 1. Example: Euclidean vs. Generalized Gini prametric ν ∗ = 3.52 [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

47 extracted references · 36 canonical work pages

  1. [9]

    H. Cover. Nearest neighbor pattern classification. IEEE transactions on information theory, 13:21–27, 1967

    work page 1967

  2. [1]

    Alfeilat, A

    H. Alfeilat, A. Hassanat, O. Lasassmeh, A. Tarawneh, M. Alhasanat, H. Eyal-Salman, and S. Prasath. Effects of distance measure choice on k-nearest neighbor classifier performance: A review. Big Data, 7(4): 221–248, 2019. doi: 10.1089/big.2018.0175

    arXiv 2019

  3. [2]

    An and S

    Q. An and S. Jiang. Sub-one quasi-norm-based k-means clustering algorithm and analyses. Neural Processing Letters, 56:175, 2024. doi: 10.1007/s11063-024-11615-y

    work page doi:10.1007/s11063-024-11615-y 2024

  4. [3]

    Arkhangelskii and L

    A. Arkhangelskii and L. Pontryagin. General Topology I, volume 17 of Encyclopaedia of Mathematical Sciences. Springer-Verlag, Berlin, 1990. ISBN 978-3-540-18178-7. doi: 10.1007/978-3-642-61255-1

    work page doi:10.1007/978-3-642-61255-1 1990

  5. [4]

    Arthur and S

    D. Arthur and S. Vassilvitskii. k-means++: The advantages of careful seeding. In Proceedings of the 18th Annual ACM-SIAM Symposium on Discrete Algorithms (SODA), pages 1027–1035. SIAM, 2007

    work page 2007

  6. [5]

    G. H. Ball and D. J. Hall. Isodata, a novel method of data analysis and pattern classification. Technical report, Stanford Research Institute, Menlo Park, CA, 1965

    work page 1965

  7. [6]

    A Scalable Algorithm for Individually Fair K-means Clustering

    M. Bateni, V . Cohen-Addad, A. Epasto, and S. Lattanzi. A scalable algorithm for individually fair k-means clustering, 2024. URL https: //arxiv.org/abs/2402.06730

    work page Pith review arXiv 2024

  8. [7]

    Carcea and R

    M. Carcea and R. Serfling. A gini autocovariance function for time series modelling. Journal of Time Series Analysis, 36(6):817–838, 2015. doi: https://doi.org/10.1111/jtsa.12130. URL https://onlinelibrary.wiley.com/ doi/abs/10.1111/jtsa.12130

    work page doi:10.1111/jtsa.12130 2015

Show all 47 references
  1. [8]

    Charpentier, S

    A. Charpentier, S. Mussard, and T. Ouraga. Principal component analysis: A generalized gini approach. European Journal of Operational Research, 294(1):236–249, 2021

    2021

  2. [10]

    T. M. Cover and P. E. Hart. Nearest neighbor pattern classification.IEEE Transactions on Information Theory, 13(1):21–27, 1967. doi: 10.1109/ TIT.1967.1053964. URL https://doi.org/10.1109/TIT.1967.1053964

    1967

  3. [11]

    X. Dang, H. Sang, and L. Weatherall. Gini covariance matrix and its affine equivariant version. Statistical Papers, 60:641–666, 2019

    2019

  4. [12]

    Derrac, S

    J. Derrac, S. García, and F. Herrera. Fuzzy nearest neighbor algorithms: Taxonomy, experimental analysis and prospects. Information Sciences, 260:98–119, 2014

    2014

  5. [13]

    C. Elkan. Using the triangle inequality to accelerate k-means. In Pro- ceedings of the 20th International Conference on Machine Learning (ICML 2003), pages 147–153, 2003. URL http://www.aaai.org/Library/ ICML/2003/icml03-022.php

    2003

  6. [14]

    Eschrich, J

    S. Eschrich, J. Ke, L. O. Hall, and D. B. Goldgof. Fast accurate fuzzy clustering through data reduction. IEEE Transactions on Fuzzy Systems, 11(2):262–270, 2003. doi: 10.1109/TFUZZ.2003.809911

    2003

  7. [15]

    Fix and J

    E. Fix and J. L. Hodges. Discriminatory analysis. nonparametric dis- crimination: Consistency properties. Technical Report 21, USAF School of Aviation Medicine, Randolph Field, Texas, 1951

    1951

  8. [16]

    C. Gini. Variabilità e mutabilità. Memorie di metodologica statistica, 3: 1–158, 1912

    1912

  9. [17]

    A. B. Hassanat. Dimensionality invariant similarity measure. Journal of American Science, 10(8):221–226, 2014

    2014

  10. [18]

    Hollander

    M. Hollander. Nonparametric statistical methods. John Wiley & Sons, 1973

    1973

  11. [19]

    P. J. Huber. Robust statistics. John Wiley & Sons, 1981

    1981

  12. [20]

    A. K. Jain. Data clustering: 50 years beyond k-means. In Joint Euro- pean Conference on Machine Learning and Knowledge Discovery in Databases, pages 3–4. Springer, 2008

    2008

  13. [21]

    A. K. Jain. Data clustering: 50 years beyond k-means. Pattern Recogni- tion Letters, 31(8):651–666, 2010. doi: 10.1016/j.patrec.2009.09.011. URL https://doi.org/10.1016/j.patrec.2009.09.011

    2010 doi

  14. [22]

    Kaufman and P

    L. Kaufman and P. J. Rousseeuw. Finding Groups in Data: An Introduc- tion to Cluster Analysis. Wiley Series in Probability and Statistics. John Wiley & Sons, Hoboken, NJ, 2005. ISBN 978-0-471-73578-5

    2005

  15. [23]

    M. G. Kendall. Rank correlation methods. Griffin, 1948

    1948

  16. [24]

    H. W. Kuhn. The hungarian method for the assignment problem. Naval Research Logistics Quarterly, 2(1-2):83–97, 1955

    1955

  17. [25]

    S. P. Lloyd. Least squares quantization in pcm. IEEE Transactions on Information Theory, 28(2):129–137, 1982. doi: 10.1109/TIT.1982. 1056489. URL https://doi.org/10.1109/TIT.1982.1056489

    1982

  18. [26]

    J. B. MacQueen. Some methods for classification and analysis of mul- tivariate observations. In Proceedings of the 5th Berkeley Symposium on Mathematical Statistics and Probability, volume 1, pages 281–297. University of California Press, 1967. URL https://projecteuclid.org/ e...

    1967

  19. [27]

    Mussard, A

    C. Mussard, A. Charpentier, and S. Mussard. Knn and k-means in gini prametric spaces, 2025. URL https://arxiv.org/abs/2501.18028

    2025 arXiv

  20. [28]

    E. U. Oti, M. O. Oberhiri-Orumah, G. Harold, and H. Chike. New k-means clustering method using minkowski’s distance as its metric. British Journal of Computer Networking and Information Technology, 4 (1):28–41, July 2021. doi: 10.52589/BJCNIT-XEPSJBWX

    2021 doi

  21. [29]

    Pelleg and A

    D. Pelleg and A. Moore. Accelerating exact k-means algorithms with geometric reasoning. In S. Chaudhuri and D. Madigan, editors, Proceed- ings of the Fifth International Conference on Knowledge Discovery and Data Mining (KDD), pages 277–281. AAAI Press, 1999

    1999

  22. [30]

    Pelleg and A

    D. Pelleg and A. Moore. X-means: Extending k-means with efficient estimation of the number of clusters. In Proceedings of the Seventeenth International Conference on Machine Learning (ICML), pages 727–734. Morgan Kaufmann, 2000

    2000

  23. [31]

    P. J. Rousseeuw. Silhouettes: a graphical aid to the interpretation and validation of cluster analysis. Journal of computational and applied mathematics, 20:53–65, 1987

    1987

  24. [32]

    Schechtman and S

    E. Schechtman and S. Yitzhaki. A measure of association based on gini’s mean difference. Communications in Statistics Theory and Methods , Series A:207–231, 1987

    1987

  25. [33]

    Schechtman and S

    E. Schechtman and S. Yitzhaki. A family of correlation coefficients based on the extended gini index. The Journal of Economic Inequality, 1:129–146, 2003

    2003

  26. [34]

    Schölkopf, A

    B. Schölkopf, A. Smola, and K.-R. Müller. Nonlinear component anal- ysis as a kernel eigenvalue problem. Neural Computation, 10(5):1299– 1319, 1998. doi: 10.1162/089976698300017467

    1998 doi

  27. [35]

    Shelef and E

    A. Shelef and E. Schechtman. A gini-based time series analysis and test for reversibility. Statistical Papers, 60(3):687–716, 2019. doi: 10.1007/s00362-016-0845-9

    2019 doi

  28. [36]

    Spearman

    C. Spearman. The proof and measurement of association between two things. The American Journal of Psychology, 15(1):72–101, 1904

    1904

  29. [37]

    Steinbach, G

    M. Steinbach, G. Karypis, and V . Kumar. A comparison of document clustering techniques. In KDD Workshop on Text Mining, 2000

    2000

  30. [38]

    Steinhaus

    H. Steinhaus. Sur la division des corps matériels en parties. Bulletin de l’Académie Polonaise des Sciences, 4(12):801–804, 1957

    1957

  31. [39]

    Syriopoulos, N

    P. Syriopoulos, N. Kalampalikis, S. Kotsiantis, et al. knn classifi- cation: a review. Annals of Mathematics and Artificial Intelligence ,

  32. [40]

    Vincent and Y

    P. Vincent and Y . Bengio. K-local hyperplane and convex distance nearest neighbor algorithms. In T. G. Dietterich, S. Becker, and Z. Ghahramani, editors, Advances in Neural Information Processing Systems, volume 14, pages 985–992, Cambridge, MA, 2001. MIT Press

    2001

  33. [41]

    Z. Wu, T. Song, and Y . Zhang. Quantum k-means algorithm based on manhattan distance. Quantum Information Processing , 21:19,

  34. [42]

    H. Xiao, K. Rasul, and R. V ollgraf. Fashion-mnist: a novel image dataset for benchmarking machine learning algorithms. arXiv preprint arXiv:1708.07747, 2017

    2017 arXiv

  35. [43]

    Yitzhaki and P

    S. Yitzhaki and P. Lambert. The relationship between the absolute deviation from a quantile and gini’s mean difference. METRON, 71: 97–104, 2013

    2013

  36. [44]

    double instrumental variable

    S. Yitzhaki and E. Schechtman. The gini instrumental variable, or the “double instrumental variable” estimator. Metron, LXII:287–313, 2004. URL ftp://metron.sta.uniroma1.it/RePEc/articoli/2004-3-287-313.pdf

    2004

  37. [45]

    Yitzhaki and E

    S. Yitzhaki and E. Schechtman. The Gini methodology: a primer on a statistical methodology, volume 272. Springer, 2013. Supplementary Materials Proof of Proposition 4 Let c(t−1) k be the centroid of group k at iteration (t − 1). If point i goes from cluster C (t−1) k at iterat...

    2013

  38. [2022]

    URL https://doi.org/10.1007/ s11128-021-03384-7

    doi: 10.1007/s11128-021-03384-7. URL https://doi.org/10.1007/ s11128-021-03384-7

    doi

  39. [2023]

    URL https://doi.org/10.1007/ s10472-023-09882-x

    doi: 10.1007/s10472-023-09882-x. URL https://doi.org/10.1007/ s10472-023-09882-x

    doi

Pith tools

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