Pith. sign in

REVIEW 3 major objections 4 minor 40 references

Number of Clusters in a Dataset: A Regularized K-means Approach

T0 review · 3 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read For regularized k-means, the penalty coefficient λ has provable bounds under an ideal-cluster model, turning an ad hoc hyperparameter into a principled interval.

desk verdict Clean geometric derivation of a principled λ interval and a parameter-free multiplicative criterion for penalized k-means under ideal clusters, but the advertised 'rigorous for all dimensions' upper bound is contradicted by the paper's own Appendix B in low dimensions — a fixable overstatement that does not sink the practical recipe. read the letter →

arxiv 2505.22991 v1 pith:M6MPXTKL submitted 2025-05-29 cs.LG cs.CV

Behzad Kamgar-Parsi , Behrooz Kamgar-Parsi This is my paper
classification cs.LGcs.CV MSC 62H3068T10
keywords regularizedk-meanspenalizednumberofclustersidealadditiveregularizationmultiplicativepenaltycoefficientboundsclustervalidation
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

Adding a penalty λk to the k-means error is a common way to let the data decide how many clusters to use, but the value of λ has been a free parameter with no principled guidance. This paper shows that if the true clusters are 'ideal'—equal non-overlapping spheres that are densely filled with data—then λ must lie strictly between Nρ²/K and NL²/(2K), where N is the number of points, K the true cluster count, ρ the half-sphere centroid offset, and L the distance between adjacent spheres. The authors further show that a multiplicative regularizer k·E_k has a natural minimum at K with no tuning parameter, and that agreeing answers from the two regularizers can resolve ambiguous cluster counts in non-ideal data. If the bounds hold, practitioners gain a concrete starting interval for λ instead of guesswork, at the cost of assuming the ideal-cluster model.

What carries the argument

The load-bearing object is the 'ideal cluster': a d-dimensional sphere of radius R and volume V, with all clusters identical, non-overlapping, and dense enough to replace sums by integrals. The argument runs on three closed-form errors—sphere E_s = VR²α, half-sphere E_h = VR²β, and dumbbell E_d = 2E_s + ½VL²—with α = d/(d+2), γ = Γ((d+2)/2)/(√π Γ((d+3)/2)), ρ = Rγ, and β = ½(α − γ²). These feed the interval Nρ²/K < λ < NL²/(2K) for the additive penalty, and also prove that k·E_k has a natural minimum at K. Two deterministic seeding algorithms (Algorithm 1 and Algorithm 2) are introduced to guarantee that, for ideal clusters, the k-means iterations reach the true centroids, so the error formulas genuinely describe the optimized configurations.

What would settle it

Compute both candidate upper bounds from the paper's formulas for, say, d = 2 and L = 2R: the perfect-dumbbell bound λ_d = ½VL² and the uneven-dumbbell bound λ_u = (V/3)(2L² − 4LRγ − R²γ²). Since inequality (B-7) holds with γ = 0.424, λ_u is strictly smaller than λ_d, which refutes the claim that the perfect-dumbbell bound is the tightest upper bound for all dimensions; a reader can verify this arithmetic in a few lines.

Watch

Extended reading notes

Core claim

Under the ideal-cluster assumption—clusters are identical, non-overlapping, densely filled d-dimensional spheres—the paper derives exact expressions for the k-means error of a sphere, a half-sphere, and a dumbbell, and uses them to prove that the additive regularized error E_k + λk has its global minimum at the correct number of clusters K exactly when Nρ²/K < λ < NL²/(2K). The lower bound comes from comparing K spheres with K+1 clusters where one sphere splits into two half-spheres; the upper bound comes from comparing K spheres with K−1 clusters where two adjacent spheres merge into a dumbbell. The paper argues that the perfect-dumbbell configuration gives the tightest upper bound in most dimensions, and recommends the midpoint λ ≈ NL²/(4K) as a repeatable choice. For the multiplicative regularizer k·E_k, no such parameter is needed: the same geometric error formulas imply the regularized error is minimized at K for all d ≥ 2 and K ≥ 2.

Load-bearing premise

The upper bound relies on the assumption that when the algorithm has one fewer cluster than the true count, the best achievable configuration is a perfect dumbbell, and that this configuration yields the tightest upper bound in every dimension; the paper's own Appendix B shows this tightness fails for d ≤ 9 when L = 2R.

Editorial extensions

If this is right

  • For ideal clusters, a user can set λ by measuring N, the smallest inter-centroid distance L, and the sphere radius R, then pick the midpoint of the interval as a repeatable default.
  • The multiplicative regularizer k·E_k requires no hyperparameter at all for ideal clusters, giving a parameter-free baseline for cluster-count selection.
  • When additive and multiplicative regularizers disagree, their consensus candidates can be treated as the plausible cluster counts, which the experiments show reduces ambiguity in overlapping or non-spherical data.
  • The same interval logic extends to other penalty shapes (logarithmic, power, exponential) by replacing the differences f(K+1)−f(K) and f(K)−f(K−1) in the bounds, as derived in Appendix C.
  • The results apply most directly in high dimensions, where cluster overlap is smaller and the ideal-cluster approximation is better.

Reading between the lines

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

  • The claim that the perfect dumbbell gives the tightest upper bound in every dimension is not supported even by the paper's own Appendix B: for L = 2R and d ≤ 9, an uneven dumbbell yields a smaller upper bound, so in those cases Eq. (11) is valid but not tight; a dimension-corrected upper bound would be needed before the interval is called rigorous in low dimensions.
  • The midpoint recommendation λ ≈ NL²/(4K) could be tested as a default on non-ideal data with known cluster structure; if it systematically under- or over-counts, the failure mode would point to how strongly the no-overlap assumption is being violated.
  • One could extend the analysis to unequal cluster radii or overlapping Gaussian clusters by numerically computing the sphere-split and dumbbell errors instead of the closed forms, retaining the same bound structure with computed quantities.
  • The consensus criterion between additive and multiplicative solutions could be formalized as a scoring rule (e.g., counting agreement across many restarts), which would quantize the ambiguity the paper reports qualitatively.
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

3 major / 4 minor

Summary. This paper studies the problem of estimating the number of clusters K in unlabeled data using regularized k-means objectives, specifically E_k^(a) = E_k + lambda k and E_k^(m) = k E_k. Under an "ideal cluster" model (equal, non-overlapping d-dimensional spheres uniformly filled with data points), the authors derive lower and upper bounds for the additive penalty coefficient lambda, stated in Eq. (11) as N rho^2 / K < lambda < N L^2 / (2K), and propose a midpoint rule combined with an assumed-K enumeration procedure. They also analyze the multiplicative regularizer, argue that it has a natural minimum at K for ideal clusters, and present experiments on simulated data, a textured image, and the Iris dataset. The central claim is that Eq. (11) provides rigorous, dimension-uniform bounds on lambda under the ideal-cluster assumption.

Significance. A principled interval for the additive penalty coefficient would address a real practical gap, since lambda is normally chosen ad hoc; the lower-bound derivation from the K versus K+1 comparison is clean, and the geometric treatment in the appendices is mostly self-contained. The multiplicative regularizer is parameter-free and appears empirically more robust, and the consensus idea between additive and multiplicative penalties is sensible. However, the paper's central rigor claim is undercut by its own Appendix B: for L = 2R and d <= 9, the uneven dumbbell gives a strictly tighter upper bound than the perfect dumbbell, so Eq. (11) is not valid for all dimensions as stated. In addition, the proposed method for selecting lambda depends on the unknown values of K and L, which limits its practical use as a principled guideline. If the upper-bound gap is repaired and the practical procedure is clarified, the contribution would be solid; in its present form the main theorem is overstated.

major comments (3)
  1. [Sec. 4 and Appendix B, Eq. (B-7)] The assertion in Sec. 4 that the perfect dumbbell "yields the tightest bound for all dimensions" is contradicted by the paper's own inequality (B-7). For non-overlapping spheres with L = 2R, inequality (B-7) holds whenever gamma > 0.243, i.e., for dimension d <= 9, and then lambda_u from Eq. (B-6) is strictly smaller than lambda_d from Eq. (B-4). For example, for d = 2, L = 2R, gamma ~= 0.424, lambda_u ~= 1.477 V R^2 while lambda_d = 2 V R^2. A value lambda in (lambda_u, lambda_d) lies inside the interval advertised by Eq. (11) yet makes E^(u)_{K-1} - E_K - lambda < 0, so K is not even a local minimum of the additively regularized error. The defense in Appendix B that lambda_d is dimension-independent and smaller for most dimensions is not a proof of tightness for all dimensions. This is a load-bearing gap in the central result, not a minor caveat; the upper bound in Eq. (11) must be replaced by a bound that is actually valid for the optimal K-1 configuration, or the claim must be restricted accordingly.
  2. [Sec. 4.1 and Eq. (12)] The proposed procedure requires K and L before lambda can be set, and L is the smallest inter-centroid distance of the unknown true configuration. The "assume K, then check" loop still needs an estimate of L for each assumed K, so the user must either supply external knowledge or estimate L from a K-cluster solution, which biases the test. The paper acknowledges this dependence, but it means that the method is not a fully data-driven principled guideline. Please specify exactly how L is estimated in the experiments and discuss how sensitive the candidate-solution sets are to that estimate.
  3. [Eq. (10) and abstract] The conditions in Eq. (10) only ensure that E^(a)_K is lower than its immediate neighbors E^(a)_{K-1} and E^(a)_{K+1}. The abstract and Sec. 4 state that the regularized error has its global minimum at K, but no argument excludes the possibility of a lower value at k < K-1 or k > K+1 under the same lambda. If the global claim is intended, a proof is needed; otherwise the text should consistently say "local minimum" rather than "global minimum."
minor comments (4)
  1. [Throughout] There are several typos in the text and figure captions, including "regularizd" in Sec. 7, "shonw" in Fig. 5, "Muliplicative" and "Clsuters" in Fig. 7, and "aditive" in Fig. 8.
  2. [Sec. 4 and Sec. 5] The citation "as may also be seen in Fig. 8" for the alpha/(2 beta) ratio is incorrect: Fig. 8 is the 8-dimensional cluster example, while the ratio is plotted in Fig. 13 of Appendix A. The cross-reference should be fixed.
  3. [Sec. 6.2] The "consensus" criterion between additive and multiplicative candidate solutions is not formally defined. Please state precisely whether it means the intersection of the two candidate sets, and how ties or near-misses are treated.
  4. [Sec. 3.1 and Sec. 3.2] The algorithms are presented as deterministic, but tie-breaking in the "choose the farthest point" steps is unspecified; for ideal equal spheres with L = 2R, ties can occur, which affects the claim that a distinct cluster is always selected as the initial centroid.

Circularity Check

1 steps flagged · score 2.0 of 10

No derivation-level circularity; one acknowledged fixed-point dependency of lambda on the unknown K keeps the paper at a low circularity score.

  1. other [Sec. 4, Eq. (11); Sec. 4.1]
    "Note that λ depends on the correct number of clusters K, the hyper-parameter we want to determine. This introduces another layer of complexity in using k-means with additive penalty."

    The interval in Eq. (11) is stated in terms of K (and L, the smallest inter-centroid distance), where K is precisely the quantity the procedure is supposed to determine. Sec. 4.1 then instructs the user to assume a value of K, select λ from the interval computed with that assumed K, run k-means, and check whether the regularized error has a minimum at the assumed K. Thus the test for K being correct uses a λ that was chosen from the hypothesis K itself, making the procedure a fixed-point or self-consistency check rather than an independent estimate of K. The paper explicitly acknowledges this dependency, labels the outputs 'potential solutions,' and does not hide the circularity.

full rationale

The core derivation of Eq. (11) is self-contained geometry: the errors E_{K-1}, E_K, and E_{K+1} are obtained from explicit integrals for spheres, half-spheres, and dumbbells in Appendix A, and the inequalities delta^{(a)}_{K-1,K} > 0 and delta^{(a)}_{K,K+1} < 0 are pure algebraic consequences of those cluster shapes. No parameter is fitted to data, and no load-bearing result is imported from the authors' prior work; the self-citation [22] is background and does not carry the argument. The only circularity-adjacent element is that the lambda interval depends on K and L, the very quantities one wants to infer, and Sec. 4.1 uses lambda obtained from an assumed K to test whether that K is a minimum. The paper acknowledges this twice, and the procedure is explicitly a search for 'potential solutions' rather than a one-shot prediction. In addition, Appendix B's inequality (B-7) shows that for d <= 9 and L = 2R the uneven dumbbell gives a smaller upper bound than the perfect dumbbell, which contradicts Sec. 4's blanket claim that the perfect dumbbell yields the tightest bound for all dimensions. That is a correctness gap in the stated rigor of the upper bound, but it is not circularity and does not raise the circularity score. Overall, the mathematical derivation is independent and self-contained; the acknowledged K-dependence of lambda warrants only a low circularity score of 2.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The central derivation rests on the ideal-cluster model (identical non-overlapping spheres, dense filling) and on a specific assumption about the optimal K-1 clustering configuration. No fitted constants enter the bounds themselves, but the midpoint lambda rule and the outlier threshold are free choices in the experimental procedure. No new physical or model entities are introduced.

free parameters (2)
  • Midpoint rule for lambda = lambda ~ N(rho^2 + 0.5 L^2)/(2K) ~ N L^2/(4K)
    The paper selects the midpoint of the derived interval to make the procedure consistent and repeatable (Eq. 12). This is a heuristic choice, not derived from optimization, and it affects whether the assumed K is accepted as a candidate solution.
  • Outlier culling threshold = not specified
    In the texture experiments, low-density points are removed using an unspecified threshold (Sec. 6.2). This threshold affects the dataset size and the clustering results, and is not quantified.
assumptions (4)
  • domain assumption Ideal clusters are d-dimensional spheres of equal radius, non-overlapping, with no outliers, and dense enough that the number of points per cluster equals the sphere volume V.
    This is the central modeling assumption stated in Sec. 3, used to replace sums with integrals throughout the derivation of E_K, E_{K-1}, E_{K+1}. It is stronger than the Gaussian assumption underlying k-means and limits the applicability of the bounds.
  • ad hoc to paper For k = K-1, the optimal clustering is either one perfect dumbbell plus K-2 spheres, or two uneven dumbbells plus K-3 spheres, and the perfect dumbbell yields the tightest upper bound for all dimensions.
    This exhaustiveness is asserted in Sec. 4 and Appendix B without a complete proof. Appendix B's own condition (B-7) shows the uneven dumbbell is tighter for d <= 9 at L = 2R, so the 'for all dimensions' part is not valid.
  • domain assumption The number of data points per ideal cluster is approximated by the sphere volume V, so N approx K V.
    The dense-cluster assumption in Sec. 3(e) lets the paper write V approx N/K in Eq. (11). This introduces an unquantified finite-N error into the bounds.
  • domain assumption Algorithms 1 and 2 are provably optimal for ideal clusters when k = K.
    Used to justify that the experimental errors match the analytic E_K for the ideal case. For k != K, optimality is not guaranteed, which affects the measured E_{K-1} and E_{K+1} used to validate the bounds.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Number of Clusters in a Dataset: A Regularized K-means Approach." pith.science (2026). https://pith.science/paper/M6MPXTKL

@misc{pith2026250522991,
  author       = {Pith},
  title        = {Pith review of: Number of Clusters in a Dataset: A Regularized K-means Approach},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/M6MPXTKL}},
  note         = {Machine review of arXiv:2505.22991}
}
abstract

Finding the number of meaningful clusters in an unlabeled dataset is important in many applications. Regularized k-means algorithm is a possible approach frequently used to find the correct number of distinct clusters in datasets. The most common formulation of the regularization function is the additive linear term $\lambda k$, where $k$ is the number of clusters and $\lambda$ a positive coefficient. Currently, there are no principled guidelines for setting a value for the critical hyperparameter $\lambda$. In this paper, we derive rigorous bounds for $\lambda$ assuming clusters are {\em ideal}. Ideal clusters (defined as $d$-dimensional spheres with identical radii) are close proxies for k-means clusters ($d$-dimensional spherically symmetric distributions with identical standard deviations). Experiments show that the k-means algorithm with additive regularizer often yields multiple solutions. Thus, we also analyze k-means algorithm with multiplicative regularizer. The consensus among k-means solutions with additive and multiplicative regularizations reduces the ambiguity of multiple solutions in certain cases. We also present selected experiments that demonstrate performance of the regularized k-means algorithms as clusters deviate from the ideal assumption.

Figures

Figures reproduced from arXiv: 2505.22991 by the authors.

Figure 1
Figure 1. Sketch of a regularized error with its global minimum at the correct number of clusters [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. A 2D examples with 10 clusters, each cluster with approximately 100 points. [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. This example is similar to Fig. 2. Clusters are the same but their locations are changed to create more [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (11 more)
Figure 4
Figure 4. Figure 4: A 2D example with 20 clusters, each cluster with approximately 200 points. [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: The clusters shonw on the left are the same as in Fig. 4, but the results are obtained by using Algorithm 2. [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: Clusters shown on the left are the same as in Fig. 5, but the inter-cluster distances are uniformly shrunk. The [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]
Figure 7
Figure 7. Figure 7: Clusters are the same as in Fig. 4, but the inter-cluster distances are uniformly shrunk. The middle figure [PITH_FULL_IMAGE:figures/full_fig_p009_7.png]
Figure 8
Figure 8. Figure 8: An example with 20 clusters in 8 dimensions. Left two figures are projections of clusters on two different 2D [PITH_FULL_IMAGE:figures/full_fig_p010_8.png]
Figure 9
Figure 9. Figure 9: Top row: A composite image with 5 Brodatz textures. Middle row: The image is represented by 1,654 points in a d = 6 dimensional space obtained from local image moments. The combined k-means algorithms with additive and multiplicative penalties yield 5 clusters (texture…
Figure 10
Figure 10. Figure 10: Plots of Iris data for all dimension pairs. Reproduced from [9] with permission. [PITH_FULL_IMAGE:figures/full_fig_p012_10.png]
Figure 11
Figure 11. Figure 11: Results of Algorithm 1 for Iris data. (Left) Additive penalty yields k = 2, 3, 4, 5, 7, 8 as solution candidates. (Right) Multiplicative penalty has its minimum at k = 3, and a very shallow local minimum at k = 7. The consensus is k = 3 where both methods have a sharp…
Figure 12
Figure 12. Figure 12: Results of Algorithm 2 for Iris data. ( [PITH_FULL_IMAGE:figures/full_fig_p012_12.png]
Figure 13
Figure 13. Figure 13: (Left) The ratio γ = ρ/R as a function of dimension d, where R is the sphere radius and ρ is distance of the half-sphere centroid from its equator plane; ρ→0 as d→∞. (Right) The ratio α/2β as a function of dimension d. The ratio goes to 1 as d→∞ [PITH_FULL_IMAGE:figu…
Figure 14
Figure 14. Figure 14: A sketch of an uneven dumbbell. The within-cluster variance for the half-sphere is given by: Eh = Z half−sphere dV k r − c k 2 , (A-11) which reduces to the following, Eh = 1 2 Es − 1 2 V ρ2 = V R2 β, with β = 1 2 ( n n+2 − γ 2 ) = 1 2 (α − γ 2 ). (A-12) A.3 Dumbbell …

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

40 extracted references · 40 canonical work pages

  1. [22]

    Kamgar-Parsi and B

    B. Kamgar-Parsi and B. Kamgar-Parsi. Penalized K-Means Algorithms for Finding the Number of Clusters. Proc. 2020 International Conference on Pattern Recognition (ICPR 2020), Milan, Italy, Jan. 10-15, 2021, pp. 969-974

    work page 2020

  2. [1]

    Ahmadian, A

    S. Ahmadian, A. Norouzi-Fard, O. Svensson, J. Ward. Better Guarantees for k-Means and Euclidean k-Median by Primal-Dual Algorithms. Proc. FOCS, pp. 61-72, 2017

    work page 2017

  3. [2]

    Recovering the number of clusters in data sets with noise features using feature rescaling factors

    R.C. de Amorim and C. Henning. Recovering the number of clusters in data sets with noise features using feature rescaling factors. arXiv:1602.06989v1, Feb. 2016

    work page Pith review arXiv 2016

  4. [3]

    Anderson

    E. Anderson. The species problem in Iris. Annals of the Missouri Botanical Garden, 23(3):457–509, 1936

    work page 1936

  5. [4]

    Arthur and S

    D. Arthur and S. Vassilvitskii. k-means++: The Advantages of Careful Seeding. Proc. Annual ACM-SIAM Sympo- sium on Discrete Algorithms (SODA), pp. 1027-1035, 2007

    work page 2007

  6. [5]

    Bachem, M

    O. Bachem, M. Lucic, S.H. Hassani, and A. Krause. Fast and Provably Good Seedings for k-Means. Proc. Neural Information Processing Systems (NIPS), pp. 55-63, 2016

    work page 2016

  7. [6]

    Recombinator-k-means: An evolutionary algorithm that exploits k-means++ for recombination

    C. Baldassi. Recombinator-k-means: Enhancing k-means++ by seeding from pools of previous runs. arXiv:1905.00531v1, May 2019

    work page Pith review arXiv 1905

  8. [7]

    A. Blum, J. Hopcroft, and R. Kannan. Foundations of Data Science, Cambridge University Press, 2020

    work page 2020

Show all 40 references
  1. [8]

    Bozdogan

    H. Bozdogan. Model selection and Akaike’s information criterion (AIC): the general theory and its analytical extensions. Psychometrika, 52(3):345–370, 1987

    1987

  2. [9]

    P.J.A. Cock. https://warwick.ac.uk/fac/sci/moac/people/students /peter_cock/r/iris_plots/

  3. [10]

    R.O. Duda, P. E. Hart, and D.G. Stork. Pattern Classification. Wiley, 2003

    2003

  4. [11]

    A. Fischer. On the number of groups in clustering. Statistics and Probability Letters, 81:1771–1781, 2011

    2011

  5. [12]

    Frank and A

    A. Frank and A. Asuncion. UCI Machine Learning Repository. University of California, Irvine, School of Information and Computer Sciences, 2010. http://archive.ics.uci.edu/ml

    2010

  6. [13]

    Franti and S

    P. Franti and S. Sieranoja. How much can k-means be improved by using better initialization and repeats? Pattern Recognition, 93(9):95-112, 2019

    2019

  7. [14]

    Fu and P.O

    W. Fu and P.O. Perry. Estimating the number of clusters using cross-validation. arXiv:1702.02658v1, Feb. 2017

    2017 arXiv

  8. [15]

    Fujita, D.Y

    A. Fujita, D.Y . Takahashi, and A.G. Patriota. A non-parametric method to estimate the number of clusters. Computational Statistics and Data Analysis, 73:27-39, 2014

    2014

  9. [16]

    Gersho and R.M

    A. Gersho and R.M. Gray. Vector Quantization and Signal Compression. Kluwer Academic Publishers, 1992

    1992

  10. [17]

    Gradshteyn and I.M

    I.S. Gradshteyn and I.M. Ryzhik. Table of Integrals, Series, and Products, Academic Press, 4th edition, p. 369, 1965

    1965

  11. [18]

    Jain and R.C

    A.K. Jain and R.C. Dubes. Algorithms for Clustering Data. Prentice-Hall, 1988

    1988

  12. [19]

    A.K. Jain. Data clustering: 50 years beyond K-means. Pattern Recognition Letters, 31(8):651-666, 2010

    2010

  13. [20]

    Kamgar-Parsi, J.A

    B. Kamgar-Parsi, J.A. Gualtieri, J.E. Devaney, and B. Kamgar-Parsi. Clustering with neural networks.Biological Cybernetics, 63(3):201-208, 1990

    1990

  14. [21]

    Kamgar-Parsi and B

    B. Kamgar-Parsi and B. Kamgar-Parsi. Dynamical Stability and Parameter Selection in Neural Optimization.Proc. International Joint Conference on Neural Networks, 4:566-571, 1992

    1992

  15. [23]

    Kanungo, D.M

    T. Kanungo, D.M. Mount, N.S. Netanyahu, C.D. Piatko, R. Silverman, and A.Y . Wu. An Efficient k-Means Clustering Algorithm: Analysis and Implementation. IEEE Trans. Pattern Analysis and Machine Intelligence , 24(7):881-892, 2002

    2002

  16. [24]

    Krzanowski and Y .T

    W.J. Krzanowski and Y .T. Lai. A criterion for determining the number of clusters in a data set. Biometrics, 44:23–34, 1988

    1988

  17. [25]

    Kulis and M.I

    B. Kulis and M.I. Jordan. Revisiting k-means: New Algorithms via Bayesian Nonparametrics. Proc. International Conference on Machine Learning (ICML), Edinburgh, UK, 2012

    2012

  18. [26]

    W. Li, E.K. Ryu, S. Osher, W. Yin, and W. Gangbo. A parallel method for earth mover’s distance.UCLA Comput. Appl. Math. Pub. (CAM) Rep., 17-12, 2017

    2017

  19. [27]

    Linde, A

    Y . Linde, A. Buzo, and R. Gray. An Algorithm for Vector Quantizer Design. IEEE Trans. Communications, 28:84-95, 1980. 14 Number of Clusters in a Dataset: A Regularized K-means Approach

    1980

  20. [28]

    S. P. Lloyd. Least squares quantization in PCM. IEEE Trans. Information Theory, 28(2):129-137, 1982. Appeared first in a 1956 AT&T Bell Labs memorandum

    1982

  21. [29]

    Manning, P

    C.D. Manning, P. Raghavan, and H. Schutze.Introduction to Information retrieval. Cambridge University Press, 2008

    2008

  22. [30]

    K.P. Murphy. Machine Learning: A Probabilistic Perspective. MIT Press, 2012

    2012

  23. [31]

    K. Rose, E. Gurewitz, and G.C. Fox. Vector quantization by deterministic annealing.IEEE Trans. Information Theory, 38(4):1249-1257, 1992

    1992

  24. [32]

    Rousseeuw

    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

  25. [33]

    J.W. Sammon. A Nonlinear Mapping for Data Structure Analysis. IEEE Trans. Computers, 18(5):401–409, 1969

    1969

  26. [34]

    G. Schwarz. Estimating the dimension of a model. The Annals of Statistics, 6(2):461–464, 1978

    1978

  27. [35]

    Sugar and G.M

    C.A. Sugar and G.M. James. Finding the number of clusters in a dataset: an information-theoretic approach. Journal of the American Statistical Association, 98(463):750-763, 2003

    2003

  28. [36]

    Szczypi´nski, M

    P. Szczypi´nski, M. Strzelecki, and A. Materka. MaZda A Software for Texture Analysis. Proc. Int’l Symp. on Information Technology Convergence, ISITC 2007

    2007

  29. [37]

    Tibshirani, G

    R. Tibshirani, G. Walther, and T. Hastie. Estimating the number of clusters in a data set via the gap statistic. Journal of the Royal Statistical Society, 63(2):411-423, 2001

    2001

  30. [38]

    Vidal, Y

    R. Vidal, Y . Ma, and S. Sastry.Generalized Principal Component Analysis, Springer-Verlag, 2016

    2016

  31. [39]

    https://en.wikipedia.org/wiki/Determining_the_number_of_ clusters_in_a_data_set. 15 Number of Clusters in a Dataset: A Regularized K-means Approach Appendix A: Single-Cluster Variances In this Appendix, we calculate the volume of a sphere in then-dimensional Euclidean space, a...

  32. [40]

    (A-4) With repeated use of (A-3), Eq

    = (2n)!√π 4nn! . (A-4) With repeated use of (A-3), Eq. (A-2) simplifies to: V = π n 2 Γ(n + 2 2 ) Rn. (A-5) The within-cluster variance, or clustering error, for then-dimensional sphere is given by: Es = ∫ sphere dV ‖ x− c‖2, (A-6) and may be expressed in the spherical coordina...

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.