pith. sign in

arxiv: 2606.00233 · v1 · pith:KE5G62KOnew · submitted 2026-05-29 · 🧮 math.ST · stat.TH

Density Evolution: A Multiscale View of Density Estimation

Kisung You This is my paper

Pith reviewed 2026-06-28 19:33 UTC · model grok-4.3

classification 🧮 math.ST stat.TH
keywords density estimationscale-space methodskernel density estimationmixture modelspersistent homologyheat flowmode treesmultiscale analysis
0
0 comments X

The pith

Density estimation is best understood as paths of densities evolving across scales rather than selection of any single estimate.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper advances a complementary perspective called density evolution, in which a data set is examined via a continuous family of densities indexed by smoothing scale, diffusion time, model complexity, density level, or noise level. Under this view Gaussian kernel estimation corresponds to heat flow on the empirical measure, scale-space and mode-tree techniques track the birth and death of modes and derivative features, finite mixtures supply compressed versions of those estimates, and cluster trees with persistent homology capture changes in level-set topology. The review assembles these links, adds three elementary structural facts about mode paths, Gaussianization semigroups, and log-concavity of concentrated mixtures, and points to uses in feature-lifetime inference and connections to diffusion-based generation. A reader would care because the shift reframes the task from picking one summary to exploring the entire multiscale probability landscape.

Core claim

Density evolution refers to the study of a data set through a path of densities indexed by smoothing scale, diffusion time, model complexity, density level, or noise level. Gaussian kernel density estimation becomes heat flow from the empirical measure; scale-space methods, critical bandwidths, mode trees, and derivative-significance displays describe the evolution of modal and derivative structure; finite mixtures and mixture reduction supply compressed representations; and cluster trees together with persistent homology summarize evolving level-set topology. The review assembles these connections, discusses inference for feature lifetimes and high-dimensional issues, notes links with score

What carries the argument

Density evolution, the path of densities indexed by smoothing scale, diffusion time, model complexity, density level, or noise level, which unifies kernel estimation as heat flow, scale-space tracking of modes, mixture compression, and persistent-homology topology.

Load-bearing premise

The reviewed connections between heat flow, scale-space, mixture reduction, and persistent homology together form a coherent and useful multiscale framework for inference.

What would settle it

A controlled comparison in which analysts using the full multiscale path obtain no measurable gain in recovering true modes, cluster structure, or topological features over analysts given only the single best fixed-scale estimate would show the perspective adds no practical value.

read the original abstract

Density estimation is often presented as a choice among parametric summaries, finite mixtures, and nonparametric smoothers. This review argues for a complementary view: a data set can be studied through a path of densities indexed by smoothing scale, diffusion time, model complexity, density level, or noise level. We call this perspective density evolution. Under this lens, Gaussian kernel density estimation is heat flow from the empirical measure; scale-space methods, critical bandwidths, mode trees, and derivative-significance displays describe the evolution of modal and derivative structure; finite mixtures and mixture reduction provide compressed representations of kernel-like estimates; and cluster trees and persistent homology summarize evolving level-set topology. We review these connections and discuss inference for feature lifetimes, high-dimensional complications, and links with score-based generative diffusion. We also include three elementary structural results: nondegenerate modes move along smooth branches, a natural moment-preserving Gaussianization semigroup is forced to be Ornstein--Uhlenbeck, and shared-covariance Gaussian mixtures become log-concave once component means are sufficiently concentrated. Together, these ideas shift attention from choosing one density estimate to studying the multiscale probability landscape.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 2 minor

Summary. The manuscript proposes viewing density estimation through the lens of 'density evolution': paths of densities indexed by smoothing scale, diffusion time, model complexity, density level, or noise level. It reviews connections from Gaussian kernel density estimation as heat flow, through scale-space methods, critical bandwidths, mode trees, mixture reduction, cluster trees, and persistent homology. Three elementary structural results are included: nondegenerate modes move along smooth branches; a natural moment-preserving Gaussianization semigroup is forced to be Ornstein-Uhlenbeck; and shared-covariance Gaussian mixtures become log-concave once component means are sufficiently concentrated. The paper discusses inference for feature lifetimes, high-dimensional complications, and links to score-based generative diffusion models.

Significance. If the connections are coherently developed and the three structural results hold with elementary proofs, the work offers a unifying multiscale perspective that could integrate tools from nonparametric statistics, topological data analysis, and generative modeling. This reframing from single estimates to evolving landscapes has potential to influence research on feature persistence and high-dimensional inference.

major comments (1)
  1. [Abstract] Abstract: the three structural results are asserted as elementary contributions, yet no derivations, proofs, or counterexamples appear in the provided text. Since these results are load-bearing for the paper's claim to include new structural insights alongside the review, explicit statements (even if elementary) or precise citations to where they are established must be added.
minor comments (2)
  1. The abstract packs many technical terms (critical bandwidths, mode trees, derivative-significance displays) without brief parenthetical definitions; a short introductory paragraph expanding the motivation for the 'density evolution' framing would improve accessibility.
  2. Ensure that all cited methods (heat flow, scale-space, persistent homology, score-based diffusion) include at least one canonical reference each, even in a review format.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the constructive comment on the abstract and the presentation of the structural results. We will revise the manuscript accordingly.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the three structural results are asserted as elementary contributions, yet no derivations, proofs, or counterexamples appear in the provided text. Since these results are load-bearing for the paper's claim to include new structural insights alongside the review, explicit statements (even if elementary) or precise citations to where they are established must be added.

    Authors: We agree that the three results require explicit support. Although described as elementary, the current manuscript states them without derivations. We will add a short dedicated section (or appendix) containing the elementary proofs for (i) smooth branches of nondegenerate modes, (ii) uniqueness of the moment-preserving Gaussianization semigroup as the Ornstein–Uhlenbeck flow, and (iii) the log-concavity threshold for shared-covariance Gaussian mixtures. This addition will be placed after the review sections and before the discussion of inference and high-dimensional issues. revision: yes

Circularity Check

0 steps flagged

No significant circularity; review framing with independent elementary results

full rationale

The paper is a review that reframes existing tools (heat flow, scale-space, mixture reduction, persistent homology) as instances of density evolution and adds three explicitly labeled elementary structural results. No equations or claims in the provided abstract reduce any stated result to a fitted parameter, self-defined quantity, or load-bearing self-citation chain. The central contribution is a perspective shift rather than a new inference procedure whose validity depends on internal definitions; the listed results are presented as derivable from standard analysis and do not reference the target multiscale view as an input. This is the normal case of a self-contained review.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Based solely on the abstract; the paper invokes standard background from nonparametric statistics, heat equations, and mixture models without listing new free parameters or invented entities.

pith-pipeline@v0.9.1-grok · 5717 in / 1117 out tokens · 24385 ms · 2026-06-28T19:33:55.900415+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

300 extracted references · 219 canonical work pages · 5 internal anchors

  1. [1]

    Van Den Bosch, Jasper Jf and Golan, Tal and Peters, Benjamin and Taylor, JohnMark and Shahbazi, Mahdiyar and Lin, Baihan and Charest, Ian and Diedrichsen, Jörn and Kriegeskorte, Nikolaus and Mur, Marieke and Schütt, Heiko H , month = sep, year =. A. doi:10.7554/eLife.107828.1 , abstract =

    work page doi:10.7554/elife.107828.1

  2. [2]

    Cheng, Zixiong and Liu, Hang , year =. Robust. doi:10.48550/ARXIV.2603.07563 , abstract =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2603.07563

  3. [3]

    Journal of the Royal Statistical Society Series B: Statistical Methodology , author =

    Huber means on. Journal of the Royal Statistical Society Series B: Statistical Methodology , author =. 2026 , pages =. doi:10.1093/jrsssb/qkaf054 , language =

    work page doi:10.1093/jrsssb/qkaf054 2026

  4. [4]

    You, Kisung , year =. Finite. doi:10.48550/ARXIV.2604.24895 , urldate =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2604.24895

  5. [5]

    Michigan Mathematical Journal , author =

    A class of. Michigan Mathematical Journal , author =. 1984 , note =. doi:10.1307/mmj/1029003026 , number =

    work page doi:10.1307/mmj/1029003026 1984

  6. [6]

    , year =

    Witkin, Andrew P. , year =. Scale-space filtering , booktitle =

  7. [7]

    Annual Review of Statistics and Its Application , author =

    Topological. Annual Review of Statistics and Its Application , author =. 2018 , pages =. doi:10.1146/annurev-statistics-031017-100045 , abstract =

    work page doi:10.1146/annurev-statistics-031017-100045 2018

  8. [8]

    and Jones, M.C

    Wand, M.P. and Jones, M.C. , month = dec, year =. Kernel. doi:10.1201/b14876 , language =

    work page doi:10.1201/b14876

  9. [9]

    The Annals of Statistics , author =

    On nonparametric estimation of density level sets , volume =. The Annals of Statistics , author =. doi:10.1214/aos/1069362732 , number =

    work page doi:10.1214/aos/1069362732

  10. [10]

    Estimating coefficient distributions in random coefficient regressions

    Variable. The Annals of Statistics , author =. doi:10.1214/aos/1176348768 , number =

    work page doi:10.1214/aos/1176348768

  11. [11]

    Journal of Computational and Graphical Statistics , author =

    A. Journal of Computational and Graphical Statistics , author =. 2010 , pages =. doi:10.1198/jcgs.2009.07049 , language =

    work page doi:10.1198/jcgs.2009.07049 2010

  12. [12]

    Journal of Classification , author =

    Estimating the. Journal of Classification , author =. 2003 , pages =. doi:10.1007/s00357-003-0004-6 , number =

    work page doi:10.1007/s00357-003-0004-6 2003

  13. [13]

    The Annals of Statistics , author =

    Fully adaptive density-based clustering , volume =. The Annals of Statistics , author =. doi:10.1214/15-AOS1331 , number =

    work page doi:10.1214/15-aos1331

  14. [14]

    Information and Control , author =

    Some inequalities satisfied by the quantities of information of. Information and Control , author =. 1959 , pages =. doi:10.1016/S0019-9958(59)90348-1 , language =

    work page doi:10.1016/s0019-9958(59)90348-1 1959

  15. [15]

    Score-Based Generative Modeling through Stochastic Differential Equations

    Song, Yang and Sohl-Dickstein, Jascha and Kingma, Diederik P. and Kumar, Abhishek and Ermon, Stefano and Poole, Ben , year =. Score-. doi:10.48550/ARXIV.2011.13456 , abstract =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2011.13456 2011

  16. [16]

    Generative modeling by estimating gradients of the data distribution , abstract =

    Song, Yang and Ermon, Stefano , year =. Generative modeling by estimating gradients of the data distribution , abstract =. Proceedings of the 33rd

  17. [17]

    Sohl-Dickstein, Jascha and Weiss, Eric and Maheswaranathan, Niru and Ganguli, Surya , editor =. Deep. Proceedings of the 32nd. 2015 , pages =

    2015

  18. [18]

    Cerebral Cortex , author =

    Local-. Cerebral Cortex , author =. 2018 , pages =. doi:10.1093/cercor/bhx179 , language =

    work page doi:10.1093/cercor/bhx179 2018

  19. [19]

    , month = feb, year =

    Silverman, B.W. , month = feb, year =. Density. doi:10.1201/9781315140919 , language =

    work page doi:10.1201/9781315140919

  20. [20]

    Journal of the Royal Statistical Society Series B: Statistical Methodology , author =

    Using. Journal of the Royal Statistical Society Series B: Statistical Methodology , author =. 1981 , pages =. doi:10.1111/j.2517-6161.1981.tb01155.x , abstract =

    work page doi:10.1111/j.2517-6161.1981.tb01155.x 1981

  21. [21]

    Technometrics , author =

    From. Technometrics , author =. 2001 , pages =. doi:10.1198/004017001316975916 , language =

    work page doi:10.1198/004017001316975916 2001

  22. [22]

    , month = mar, year =

    Scott, David W. , month = mar, year =. Multivariate. doi:10.1002/9781118575574 , language =

    work page doi:10.1002/9781118575574

  23. [23]

    Journal d'Analyse Mathématique , author =

    On. Journal d'Analyse Mathématique , author =. 1951 , pages =. doi:10.1007/BF02790092 , language =

    work page doi:10.1007/bf02790092 1951

  24. [24]

    The Annals of Statistics , author =

    Asymptotics and optimal bandwidth selection for highest density region estimation , volume =. The Annals of Statistics , author =. doi:10.1214/09-AOS766 , number =

    work page doi:10.1214/09-aos766

  25. [25]

    The Annals of Mathematical Statistics , author =

    Remarks on. The Annals of Mathematical Statistics , author =. 1956 , pages =. doi:10.1214/aoms/1177728190 , language =

    work page doi:10.1214/aoms/1177728190 1956

  26. [26]

    The Annals of Statistics , author =

    Generalized density clustering , volume =. The Annals of Statistics , author =. doi:10.1214/10-AOS797 , number =

    work page doi:10.1214/10-aos797

  27. [27]

    Bernoulli , author =

    Optimal rates for plug-in estimators of density level sets , volume =. Bernoulli , author =. doi:10.3150/09-BEJ184 , number =

    work page doi:10.3150/09-bej184

  28. [28]

    Journal of the Royal Statistical Society Series B: Statistical Methodology , author =

    On. Journal of the Royal Statistical Society Series B: Statistical Methodology , author =. 1997 , pages =. doi:10.1111/1467-9868.00095 , abstract =

    work page doi:10.1111/1467-9868.00095 1997

  29. [29]

    The Annals of Statistics , author =

    The topography of multivariate normal mixtures , volume =. The Annals of Statistics , author =. doi:10.1214/009053605000000417 , number =

    work page doi:10.1214/009053605000000417

  30. [30]

    Computational Statistics & Data Analysis , author =

    Alternating kernel and mixture density estimates , volume =. Computational Statistics & Data Analysis , author =. 2000 , pages =. doi:10.1016/S0167-9473(00)00003-7 , language =

    work page doi:10.1016/s0167-9473(00)00003-7 2000

  31. [31]

    The Annals of Statistics , author =

    Measuring. The Annals of Statistics , author =. doi:10.1214/aos/1176324626 , number =

    work page doi:10.1214/aos/1176324626

  32. [32]

    and Wang, Bei and Zheng, Yan , editor =

    Phillips, Jeff M. and Wang, Bei and Zheng, Yan , editor =. Geometric. 2015 , pages =. doi:10.4230/LIPICS.SOCG.2015.857 , booktitle =

    work page doi:10.4230/lipics.socg.2015.857 2015

  33. [33]

    The Annals of Mathematical Statistics , author =

    On. The Annals of Mathematical Statistics , author =. 1962 , pages =. doi:10.1214/aoms/1177704472 , language =

    work page doi:10.1214/aoms/1177704472 1962

  34. [34]

    Journal of Computational and Graphical Statistics , author =

    The. Journal of Computational and Graphical Statistics , author =. 1993 , pages =. doi:10.1080/10618600.1993.10474599 , language =

    work page doi:10.1080/10618600.1993.10474599 1993

  35. [35]

    The Annals of Statistics , author =

    Nonparametric testing of the existence of modes , volume =. The Annals of Statistics , author =. doi:10.1214/aos/1031594735 , number =

    work page doi:10.1214/aos/1031594735

  36. [36]

    , year =

    Lindsay, Bruce G. , year =. Mixture models: theory, geometry, and applications , isbn =

  37. [37]

    A Universal Prior for Integers and Estimation by Minimum Description Length

    On a. The Annals of Statistics , author =. doi:10.1214/aos/1176346412 , number =

    work page doi:10.1214/aos/1176346412

  38. [38]

    Journal of Applied Statistics , author =

    Scale-space theory: a basic tool for analyzing structures at different scales , volume =. Journal of Applied Statistics , author =. 1994 , pages =. doi:10.1080/757582976 , language =

    work page doi:10.1080/757582976 1994

  39. [39]

    Biological Cybernetics , author =

    The structure of images , volume =. Biological Cybernetics , author =. 1984 , pages =. doi:10.1007/BF00336961 , language =

    work page doi:10.1007/bf00336961 1984

  40. [40]

    The Annals of Mathematical Statistics , author =

    Consistency of the. The Annals of Mathematical Statistics , author =. 1956 , pages =. doi:10.1214/aoms/1177728066 , language =

    work page doi:10.1214/aoms/1177728066 1956

  41. [41]

    Jordan, D

    The. SIAM Journal on Mathematical Analysis , author =. 1998 , pages =. doi:10.1137/S0036141096303359 , language =

    work page doi:10.1137/s0036141096303359 1998

  42. [42]

    Denoising

    Ho, Jonathan and Jain, Ajay and Abbeel, Pieter , editor =. Denoising. Advances in. 2020 , pages =

    2020

  43. [43]

    , year =

    Hartigan, John A. , year =. Clustering algorithms , isbn =

  44. [44]

    Hall, Peter and York, Matthew , year =. On the. Statistica Sinica , publisher =

  45. [45]

    The Annals of Statistics , author =

    Confidence bands in density estimation , volume =. The Annals of Statistics , author =. doi:10.1214/09-AOS738 , number =

    work page doi:10.1214/09-aos738

  46. [46]

    Journal of the Royal Statistical Society Series B: Statistical Methodology , author =

    Non-. Journal of the Royal Statistical Society Series B: Statistical Methodology , author =. 2016 , pages =. doi:10.1111/rssb.12111 , abstract =

    work page doi:10.1111/rssb.12111 2016

  47. [47]

    The Annals of Statistics , author =

    Nonparametric ridge estimation , volume =. The Annals of Statistics , author =. doi:10.1214/14-AOS1218 , number =

    work page doi:10.1214/14-aos1218

  48. [48]

    IEEE Transactions on Information Theory , author =

    The estimation of the gradient of a density function, with applications in pattern recognition , volume =. IEEE Transactions on Information Theory , author =. 1975 , pages =. doi:10.1109/TIT.1975.1055330 , number =

    work page doi:10.1109/tit.1975.1055330 1975

  49. [49]

    Journal of the American Statistical Association , author =

    Model-. Journal of the American Statistical Association , author =. 2002 , pages =. doi:10.1198/016214502760047131 , language =

    work page doi:10.1198/016214502760047131 2002

  50. [50]

    Biometrika , author =

    Mode testing via the excess mass estimate , volume =. Biometrika , author =. 2001 , pages =. doi:10.1093/biomet/88.2.499 , language =

    work page doi:10.1093/biomet/88.2.499 2001

  51. [51]

    IEEE Transactions on Pattern Analysis and Machine Intelligence , author =

    Unsupervised. IEEE Transactions on Pattern Analysis and Machine Intelligence , author =. 2002 , pages =. doi:10.1109/34.990138 , number =

    work page doi:10.1109/34.990138 2002

  52. [52]

    Optimal Stopping and Sequential Tests which Minimize the Maximum Expected Sample Size

    A. The Annals of Statistics , author =. doi:10.1214/aos/1176342360 , number =

    work page doi:10.1214/aos/1176342360

  53. [53]

    The Annals of Statistics , author =

    Confidence sets for persistence diagrams , volume =. The Annals of Statistics , author =. doi:10.1214/14-AOS1252 , number =

    work page doi:10.1214/14-aos1252

  54. [54]

    E., & Raftery, A

    Bayesian. Journal of the American Statistical Association , author =. 1995 , pages =. doi:10.1080/01621459.1995.10476550 , language =

    work page doi:10.1080/01621459.1995.10476550 1995

  55. [55]

    Eldridge, Justin and Belkin, Mikhail and Wang, Yusu , editor =. Beyond. Proceedings of. 2015 , pages =

    2015

  56. [56]

    , month = may, year =

    Efron, Bradley and Tibshirani, R.J. , month = may, year =. An. doi:10.1201/9780429246593 , language =

    work page doi:10.1201/9780429246593

  57. [57]

    The Annals of Statistics7(1), 1 – 26 (1979).https://doi.org/10.1214/aos/1176344552,https: //doi.org/10.1214/aos/11763445525

    Bootstrap. The Annals of Statistics , author =. doi:10.1214/aos/1176344552 , number =

    work page doi:10.1214/aos/1176344552

  58. [58]

    , year =

    Edelsbrunner, Herbert and Harer, J. , year =. Computational

  59. [59]

    Computational Statistics & Data Analysis , author =

    Feature significance for multivariate kernel density estimation , volume =. Computational Statistics & Data Analysis , author =. 2008 , pages =. doi:10.1016/j.csda.2008.02.035 , language =

    work page doi:10.1016/j.csda.2008.02.035 2008

  60. [60]

    and Thomas, Joy A

    Cover, Thomas M. and Thomas, Joy A. , month = sep, year =. Elements of. doi:10.1002/047174882X , language =

    work page doi:10.1002/047174882x

  61. [61]

    IEEE Transactions on Pattern Analysis and Machine Intelligence , author =

    Mean shift: a robust approach toward feature space analysis , volume =. IEEE Transactions on Pattern Analysis and Machine Intelligence , author =. 2002 , pages =. doi:10.1109/34.1000236 , number =

    work page doi:10.1109/34.1000236 2002

  62. [62]

    IEEE Transactions on Pattern Analysis and Machine Intelligence , author =

    Mean shift, mode seeking, and clustering , volume =. IEEE Transactions on Pattern Analysis and Machine Intelligence , author =. 1995 , pages =. doi:10.1109/34.400568 , number =

    work page doi:10.1109/34.400568 1995

  63. [63]

    The Annals of Statistics , author =

    Asymptotic theory for density ridges , volume =. The Annals of Statistics , author =. doi:10.1214/15-AOS1329 , number =

    work page doi:10.1214/15-aos1329

  64. [64]

    Journal of Machine Learning Research , author =

    Robust. Journal of Machine Learning Research , author =. 2018 , pages =

    2018

  65. [65]

    Foundations of Computational Mathematics , author =

    Geometric. Foundations of Computational Mathematics , author =. 2011 , pages =. doi:10.1007/s10208-011-9098-0 , language =

    work page doi:10.1007/s10208-011-9098-0 2011

  66. [66]

    The Annals of Statistics , author =

    Scale space view of curve estimation , volume =. The Annals of Statistics , author =. doi:10.1214/aos/1016218224 , number =

    work page doi:10.1214/aos/1016218224

  67. [67]

    and Gray, Robert J

    Journal of the American Statistical Association , author =. 1999 , pages =. doi:10.1080/01621459.1999.10474186 , language =

    work page doi:10.1080/01621459.1999.10474186 1999

  68. [68]

    IEEE Transactions on Information Theory , author =

    Consistent. IEEE Transactions on Information Theory , author =. 2014 , pages =. doi:10.1109/TIT.2014.2361055 , number =

    work page doi:10.1109/tit.2014.2361055 2014

  69. [69]

    Rates of convergence for the cluster tree , booktitle =

    Chaudhuri, Kamalika and Dasgupta, Sanjoy , year =. Rates of convergence for the cluster tree , booktitle =

  70. [70]

    Electronic Journal of Statistics , author =

    Data-driven density derivative estimation, with applications to nonparametric clustering and bump hunting , volume =. Electronic Journal of Statistics , author =. doi:10.1214/13-EJS781 , number =

    work page doi:10.1214/13-ejs781

  71. [71]

    Carreira-Perpi. On the. Scale. 2003 , pages =. doi:10.1007/3-540-44935-3_44 , language =

    work page doi:10.1007/3-540-44935-3_44 2003

  72. [72]

    Bulletin of the American Mathematical Society , author =

    Topology and data , volume =. Bulletin of the American Mathematical Society , author =. 2009 , pages =. doi:10.1090/S0273-0979-09-01249-X , language =

    work page doi:10.1090/s0273-0979-09-01249-x 2009

  73. [73]

    ACM Transactions on Knowledge Discovery from Data , author =

    Hierarchical. ACM Transactions on Knowledge Discovery from Data , author =. 2015 , pages =. doi:10.1145/2733381 , abstract =

    work page doi:10.1145/2733381 2015

  74. [74]

    The Annals of Statistics , author =

    Kernel density estimation via diffusion , volume =. The Annals of Statistics , author =. doi:10.1214/10-AOS799 , number =

    work page doi:10.1214/10-aos799

  75. [75]

    SIAM Journal on Mathematics of Data Science , author =

    Kernel. SIAM Journal on Mathematics of Data Science , author =. 2026 , pages =. doi:10.1137/24M1703677 , language =

    work page doi:10.1137/24m1703677 2026

  76. [76]

    Optimal Stopping and Sequential Tests which Minimize the Maximum Expected Sample Size

    On. The Annals of Statistics , author =. doi:10.1214/aos/1176342558 , number =

    work page doi:10.1214/aos/1176342558

  77. [77]

    2014 , PAGES =

    Bakry, Dominique and Gentil, Ivan and Ledoux, Michel , year =. Analysis and. doi:10.1007/978-3-319-00227-9 , language =

    work page doi:10.1007/978-3-319-00227-9

  78. [78]

    IEEE Transactions on Pattern Analysis and Machine Intelligence , author =

    Uniqueness of the. IEEE Transactions on Pattern Analysis and Machine Intelligence , author =. 1986 , pages =. doi:10.1109/TPAMI.1986.4767749 , number =

    work page doi:10.1109/tpami.1986.4767749 1986

  79. [79]

    Statistics and Computing , author =

    Clustering via nonparametric density estimation , volume =. Statistics and Computing , author =. 2007 , pages =. doi:10.1007/s11222-006-9010-y , language =

    work page doi:10.1007/s11222-006-9010-y 2007

  80. [80]

    1982 , issn =

    Reverse-time diffusion equation models , volume =. Stochastic Processes and their Applications , author =. 1982 , pages =. doi:10.1016/0304-4149(82)90051-5 , language =

    work page doi:10.1016/0304-4149(82)90051-5 1982

Showing first 80 references.