Rate of convergence for geometric inference based on the empirical Christoffel function
ESAIM: Probability and Statistics, Tome 26 (2022), pp. 171-207

We consider the problem of estimating the support of a measure from a finite, independent, sample. The estimators which are considered are constructed based on the empirical Christoffel function. Such estimators have been proposed for the problem of set estimation with heuristic justifications. We carry out a detailed finite sample analysis, that allows us to select the threshold and degree parameters as a function of the sample size. We provide a convergence rate analysis of the resulting support estimation procedure. Our analysis establishes that we may obtain finite sample bounds which are comparable to existing rates for different set estimation procedures. Our results rely on concentration inequalities for the empirical Christoffel function and on estimates of the supremum of the Christoffel-Darboux kernel on sets with smooth boundaries, that can be considered of independent interest.

DOI : 10.1051/ps/2022003
Classification : 62G05, 42C05
Keywords: Support estimation, christoffel function, concentration, finite sample, convergence rate
@article{PS_2022__26_1_171_0,
     author = {Vu, Mai Trang and Bachoc, Fran\c{c}ois and Pauwels, Edouard},
     title = {Rate of convergence for geometric inference based on the empirical {Christoffel} function},
     journal = {ESAIM: Probability and Statistics},
     pages = {171--207},
     year = {2022},
     publisher = {EDP-Sciences},
     volume = {26},
     doi = {10.1051/ps/2022003},
     mrnumber = {4387184},
     language = {en},
     url = {https://www.numdam.org/articles/10.1051/ps/2022003/}
}
TY  - JOUR
AU  - Vu, Mai Trang
AU  - Bachoc, François
AU  - Pauwels, Edouard
TI  - Rate of convergence for geometric inference based on the empirical Christoffel function
JO  - ESAIM: Probability and Statistics
PY  - 2022
SP  - 171
EP  - 207
VL  - 26
PB  - EDP-Sciences
UR  - https://www.numdam.org/articles/10.1051/ps/2022003/
DO  - 10.1051/ps/2022003
LA  - en
ID  - PS_2022__26_1_171_0
ER  - 
%0 Journal Article
%A Vu, Mai Trang
%A Bachoc, François
%A Pauwels, Edouard
%T Rate of convergence for geometric inference based on the empirical Christoffel function
%J ESAIM: Probability and Statistics
%D 2022
%P 171-207
%V 26
%I EDP-Sciences
%U https://www.numdam.org/articles/10.1051/ps/2022003/
%R 10.1051/ps/2022003
%G en
%F PS_2022__26_1_171_0
Vu, Mai Trang; Bachoc, François; Pauwels, Edouard. Rate of convergence for geometric inference based on the empirical Christoffel function. ESAIM: Probability and Statistics, Tome 26 (2022), pp. 171-207. doi: 10.1051/ps/2022003

[1] C. Aaron and O. Bodart, Local convex hull support and boundary estimation. J. Multivar. Anal. 147 (2016) 82–101. | MR | DOI

[2] C.C. Aggarwal and S. Sathe, Theoretical foundations and algorithms for outlier ensembles. ACM Sigkdd Explor. Newsl. 17 (2015) 24–47. | DOI

[3] N. Aronszajn, Theory of reproducing kernels. Trans. Am. Math. Soc. 68 (1950) 337–404. | MR | Zbl | DOI

[4] M. Berger and B. Gostiaux, Differential Geometry: Manifolds, Curves, and Surfaces: Manifolds, Curves, and Surfaces, Vol. 115, Springer Science & Business Media (2012). | Zbl

[5] A. Berlinet and C. Thomas-Agnan, Reproducing kernel Hilbert spaces in probability and statistics. Springer Science & Business Media (2011). | MR | Zbl

[6] R. Berman, S. Boucksom and D.W. Nyström, Fekete points and convergence towards equilibrium measures on complex manifolds. Acta Math. 207 (2011) 1–27. | MR | Zbl | DOI

[7] G. Biau, B. Cadre and B. Pelletier, Exact rates in density support estimation. J. Multivar. Anal. 99 (2008) 2185–2207. | MR | Zbl | DOI

[8] L. Bos, Asymptotics for the Christoffel function for Jacobi like weights on a ball in 𝐑 m . New Zealand J. Math. 23 (1994) 109. | MR | Zbl

[9] L. Bos, B. Della Vecchia and G. Mastroianni, On the asymptotics of Christoffel functions for centrally symmetric weight functions on the ball in d . Rend. Cir. Mat. 52 (1998) 277–290. | MR | Zbl

[10] J. Chevalier, Estimation du support et du contour du support d’une loi de probabilité. Ann. Inst. Henri Poincaré Stat. 12 (1976) 339–364. | MR | Zbl | Numdam

[11] A. Cholaquidis, A. Cuevas and R. Fraiman, On Poincaré cone property. Ann. Stat. 42 (2014) 255–284. | MR | Zbl | DOI

[12] A. Cuevas and R. Fraiman, A plug-in approach to support estimation. Ann. Stat. 25 (1997) 2300–2312. | MR | Zbl | DOI

[13] A. Cuevas, R. Fraiman and B. Pateiro-López, On statistical properties of sets fulfilling rolling-type conditions. Adv. Appl. Prob. 44 (2012) 311–329. | MR | Zbl | DOI

[14] A. Cuevas, W. González-Manteiga and A. Rodríguez-Casal, Plug-in estimation of general level sets. Aust. New Zealand J. Stat. 48 (2006) 7–19. | MR | Zbl | DOI

[15] A. Cuevas and A. Rodríguez-Casal, On boundary estimation. Adv. Appl. Prob. 36 (2004) 340–354. | MR | Zbl | DOI

[16] L. Devroye and G.L. Wise, Detection of abnormal behavior via nonparametric estimation of the support. SIAM J. Appl. Math. 38 (1980) 480–488. | MR | Zbl | DOI

[17] D. Dua and C. Graff, UCI Machine Learning Repository (2017).

[18] C.F. Dunkl and Y. Xu, Vol. 155 of Orthogonal polynomials of several variables. Cambridge University Press (2014). | MR | Zbl | DOI

[19] H. Edelsbrunner, D. Kirkpatrick and R. Seidel, On the shape of a set of points in the plane. IEEE Trans. Inf. Theory 29 (1983) 551–559. | MR | Zbl | DOI

[20] J. Geffroy, Sur un probleme d’estimation géométrique. Publ. Inst. Statist. Univ. Paris 13 (1964) 191–210. | MR | Zbl

[21] F. Keller, E. Muller and K. Bohm, HiCS: High contrast subspaces for density-based outlier ranking, in 2012 IEEE 28th international conference on data engineering, IEEE (2012) 1037–1048.

[22] A. Kroó and D. Lubinsky, Christoffel functions and universality in the bulk for multivariate orthogonal polynomials. Can. J. Math. 65 (2013) 600–620. | MR | Zbl | DOI

[23] A. Kroó and D. Lubinsky, Christoffel functions and universality on the boundary of the ball. Acta Math. Hung. 140 (2013) 117–133. | MR | DOI

[24] J.B. Lasserre and E. Pauwels, The empirical Christoffel function with applications in data analysis. Adv. Comput. Math. 45 (2019) 1439–1468. | MR | DOI

[25] E. Mammen and A.B. Tsybakov, Asymptotical minimax recovery of sets with smooth boundaries. Ann. Stat. 23 (1995) 502–524. | MR | Zbl | DOI

[26] S. Marx, E. Pauwels, T. Weisser, D. Henrion and J. Lasserre, Tractable semi-algebraic approximation using Christoffel-Darboux kernel. Preprint (2019). | arXiv | MR

[27] I.S. Molchanov, A limit theorem for solutions of inequalities. Scand. J. Stat. 25 (1998) 235–242. | MR | Zbl | DOI

[28] T. Patschkowski and A. Rohde, Adaptation to lowest density regions with application to support recovery. Ann. Stat. 44 (2016) 255–287. | MR | DOI

[29] E. Pauwels and J.B. Lasserre, Sorting out typicality with the inverse moment matrix SOS polynomial, in Advances in Neural Information Processing Systems (2016) 190–198.

[30] E. Pauwels, M. Putinar and J.-B. Lasserre, Data analysis from empirical moments and the Christoffel function. Found. Comput. Math. 21 (2021) 243–273. | MR | DOI

[31] F. Piazzon, Bernstein Markov properties and applications, Ph.D. thesis, Dipartimento di Matematica, Università degli Studi di Padova (2016).

[32] W. Polonik, Measuring mass concentrations and estimating density contour clusters-an excess mass approach. Ann. Stat. 23 (1995) 855–881. | MR | Zbl | DOI

[33] A. Rényi and R. Sulanke, Über die konvexe Hülle von n zufällig gewählten Punkten. Prob. Theory Related Fields 2 (1963) 75–84. | MR | Zbl

[34] P. Rigollet and R. Vert, Optimal rates for plug-in estimators of density level sets. Bernoulli 15 (2009) 1154–1178. | MR | Zbl | DOI

[35] A. Rodríguez Casal Set estimation under convexity type assumptions. Ann. Inst. Henri Poincaré, Prob. Stat. 43 (2007) 763–774. | MR | Zbl | Numdam | DOI

[36] H.L. Royden and P. Fitzpatrick, Real analysis, vol. 32. Macmillan New York (1988). | MR

[37] A. Singh, C. Scott and R. Nowak, Adaptive Hausdorff estimation of density level sets. Ann. Stat. 37 (2009) 2760–2782. | MR | Zbl | DOI

[38] G. Szegö, Vol. 23 of Orthogonal polynomials. American Mathematical Soc. (1939). | MR | Zbl

[39] V. Totik, Asymptotics for Christoffel functions for general measures on the real line. J. d’Anal. Math. 81 (2000) 283–303. | MR | Zbl | DOI

[40] A.B. Tsybakov, On nonparametric estimation of density level sets. Ann. Stat. 25 (1997) 948–969. | MR | Zbl | DOI

[41] R. Vershynin, Introduction to the non-asymptotic analysis of random matrices. Preprint (2010). | arXiv | MR

[42] G. Walther, Granulometric smoothing. Ann. Stat. 25 (1997) 2273–2299. | MR | Zbl | DOI

[43] G. Walther, On a generalization of Blaschke’s rolling theorem and the smoothing of surfaces. Math. Methods Appl. Sci. 22 (1999) 301–316. | MR | Zbl | DOI

[44] Y. Xu, Asymptotics of the Christoffel functions on a simplex in 𝐑 d . J. Approx. Theory 99 (1999) 122–133. | MR | Zbl | DOI

[45] Y. Xu, Summability of Fourier orthogonal series for Jacobi weight on a ball in 𝐑 d . Trans. Am. Math. Soc. 351 (1999) 2439–2458. | MR | Zbl | DOI

Cité par Sources :