Adaptive detection of functions of large number of variables

Yu. I. Ingster   ; I. A. Suslina

Zapiski Nauchnykh Seminarov POMI, Probability and statistics. Part 15, Tome 368 (2009), pp. 156-170

Résumé

A major difficulty arising in statistics of multi-variable functions is “the curse of dimensionality”: the rates of accuracy in estimation and separation rates in detection problems behave poorly when the number of variables increases. This difficulty arises for most popular functional classes such as Sobolev or Hölder balls. In the paper [9], it was considered functional classes of a new type, first introduced by Sloan and Wożniakowski in [14]. These classes are the balls $\mathcal F_{\sigma,s}$ in the weighed tensor product spaces that are characterized by two parameters: $\sigma>0$ is a “smoothness” parameter, and $s>0$ determines the weight sequence which characterizes “importance” of the variables. In particular, it was shown in [9] that under the white Gaussian noise model, the log-asymptotics of separation rates in detection are similar to those for one-variable functions of the smoothness $\sigma^*=\min(s,\sigma)$ independently of the original problem dimensions; thus the curse of dimensionality is “lifted.” However the test procedure depends on parameters $(\sigma,s)$ which are unknown typically. In this paper, we propose a common test procedure that does not depend on parameters $(\sigma,s)$ and provides the same log-asymptotics of separation rates uniformly over any compact set of parameters $(\sigma,s)$. Also we give independent simple proof of the log-asymptotics of separation rates in the problem. Bibl. – 16 titles.

Bibliographie

[1] M. S. Ermakov, “Minimaksnoe obnaruzhenie signala v gaussovskom belom shume”, Teoriya veroyatn. ee primen., 35:4 (1990), 704–715 | MR | Zbl

[2] I. A. Ibragimov, R. Z. Khasminskii, “Some estimation problems on Infinite dimensional Gaussian white noise”, Festschrift for Lusien Le Cam, Research papers in Probability and Statictics, Springer-Verlag, NY, 1997, 259–274 | DOI | Zbl

[3] Yu. I. Ingster, “O minimaksnom neparametricheskom obnaruzhenii signala v gaussovskom belom shume”, Problemy peredachi informatsii, 18:2 (1982), 61–73 | MR | Zbl

[4] Yu. I. Ingster, “Asymptotically minimax testing of nonparametric hypotheses”, Proc. 4th Vilnius Conference on Probab. Theory and Math. Stat., Vol. 1, VNU Science Press, 1987, 553–573 | MR

[5] Yu. I. Ingster, “Asymptotically minimax hypothesis testing for nonparametric alternatives. I”, Math. Methods Statist., 2 (1993), 85–114 ; “II”, 171–189 ; “III”, 249–268 | MR | Zbl | MR | Zbl | MR | Zbl

[6] Yu. I. Ingster, I. A. Suslina, Nonparametric Goodness-of-Fit Testing under Gaussian Model, Lect Notes Statist., 169, Springer-Verlag, New York, 2002 | MR

[7] Yu. I. Ingster, I. A. Suslina, “Nonparametric hypothesis testing for small type I errors. I”, Math. Methods Statist., 13 (2004), 409–459 | MR | Zbl

[8] Yu. I. Ingster, I. A. Suslina, “On estimation and detection of smooth function of many variables”, Math. Methods Statist., 14 (2005), 299–331 | MR

[9] Yu. I. Ingster, I. A. Suslina, “Estimation and detection of high-variable function from Sloan–Wożniakowski space”, Math. Methods Statist., 16 (2007), 318–353 | DOI | MR | Zbl

[10] Yu. I. Ingster, I. A. Suslina, “Otsenivanie i proverka gipotez dlya funktsii iz tenzornykh proizvedenii prostranstv”, Zap. nauchn. semin. POMI, 351, 2007, 180–218 | MR

[11] F. Y. Kuo, J. H. Sloan, “Lifting the curse of dimensionality”, Notices Amer. Math. Soc., 52 (2005), 1320–1329 | MR | Zbl

[12] Y. Lin, “Tensor product space ANOVA model”, Ann. Statist., 28 (2000), 734–755 | DOI | MR | Zbl

[13] A. V. Skorokhod, Integrirovanie v gilbertovom prostranstve, Nauka, M., 1975

[14] I. H. Sloan, H. Wożniakowski, “When are quazi-Monte Carlo algorithms efficient for high dimensional integrals?”, J. Complexity, 14 (1998), 1–33 | DOI | MR | Zbl

[15] V. G. Spokoiny, “Adaptive hypothesis testing using wavelets”, Ann. Statist., 24 (1996), 2477–2498 | DOI | MR | Zbl

[16] H. Wożniakowski, “Tractability of multivariate problems for weighted spaces of functions”, Appoximation and Probability, Banach Center Publications, 72, Institute of Mathematics, Polish Academy of Science, Warszawa, 2006 | MR