Abstract
This paper addresses the problem of constructing and analyzing estimators for the regression problem in supervised learning. Recently, there has been great interest in studying universal estimators. The term “universal” means that, on the one hand, the estimator does not depend on the a priori assumption that the regression function f ρ belongs to some class F from a collection of classes F and, on the other hand, the estimation error for f ρ is close to the optimal error for the class F. This paper is an illustration of how the general technique of constructing universal estimators, developed in the author’s previous paper, can be applied in concrete situations. The setting of the problem studied in the paper has been motivated by a recent paper by Smale and Zhou. The starting point for us is a kernel K(x, u) defined on X × Ω. On the base of this kernel, we build an estimator that is universal for classes defined in terms of nonlinear approximations with regard to the system {K(·, u)} uεΩ. To construct an easily implementable estimator, we apply the relaxed greedy algorithm.
Similar content being viewed by others
References
A. R. Barron, “Universal Approximation Bounds for Superposition of n Sigmoidal Functions,” IEEE Trans. Inf. Theory 39(3), 930–945 (1993).
P. Binev, A. Cohen, W. Dahmen, R. DeVore, and V. Temlyakov, “Universal Algorithms for Learning Theory. Part I: Piecewise Constant Functions,” J. Mach. Learn. Res. 6, 1297–1321 (2005).
B. Carl, “Entropy Numbers, s-Numbers, and Eigenvalue Problems,” J. Funct. Anal. 41, 290–306 (1981).
A. Barron, A. Cohen, W. Dahmen, and R. DeVore, “Approximation and Learning by Greedy Algorithms,” Manuscript (2005), http://www.ann.jussieu.fr/:_cohen/greedy.pdf.gz
F. Cucker and S. Smale, “On the Mathematical Foundations of Learning,” Bull. Am. Math. Soc. 39, 1–49 (2002).
R. DeVore, G. Kerkyacharian, D. Picard, and V. Temlyakov, “On Mathematical Methods of Learning,” Ind. Math. Inst. Res. Rep. No. 10 (Univ. South Carolina, Columbia, 2004).
R. DeVore, G. Kerkyacharian, D. Picard, and V. Temlyakov, “Mathematical Methods for Supervised Learning,” Ind. Math. Inst. Res. Rep. No. 22 (Univ. South Carolina, Columbia, 2004); “Approximation Methods for Supervised Learning,” Found. Comput. Math. 6, 3–58 (2006).
L. Györfi, M. Kohler, A. Krzyzak, and H. Walk, A Distribution-Free Theory of Nonparametric Regression (Springer, Berlin, 2002).
R. A. DeVore and V. N. Temlyakov, “Some Remarks on Greedy Algorithms,” Adv. Comput. Math. 5, 173–187 (1996).
P. J. Huber, “Projection Pursuit,” Ann. Stat. 13, 435–475 (1985).
L. Jones, “On a Conjecture of Huber Concerning the Convergence of Projection Pursuit Regression,” Ann. Stat. 15, 880–882 (1987).
G. Kerkyacharian and D. Picard, “Thresholding in Learning Theory,” math.ST/0510271.
S. V. Konyagin and V. N. Temlyakov, “The Entropy in the Learning Theory. Error Estimates,” Ind. Math. Inst. Res. Rep. No. 09 (Univ. South Carolina, Columbia, 2004); Constr. Approx. 25, 1–27 (2007).
W. S. Lee, P. L. Bartlett, and R. C. Williamson, “Efficient Agnostic Learning of Neural Networks with Bounded Fan-in,” IEEE Trans. Inf. Theory 42(6), 2118–2132 (1996).
W. S. Lee, P. L. Bartlett, and R. C. Williamson, “The Importance of Convexity in Learning with Squared Loss,” IEEE Trans. Inf. Theory 44(5), 1974–1980 (1998).
E. Schmidt, “Zur Theorie der linearen und nichtlinearen Integralgleichungen. I,” Math. Ann. 63, 433–476 (1906–1907).
S. Smale and D.-X. Zhou, “Learning Theory Estimates via Integral Operators and Their Approximations,” Manuscript (2005), http://www.tti-c.org/smale_papers/sampIII5412.pdf
V. N. Temlyakov, “Optimal Estimators in Learning Theory,” Ind. Math. Inst. Res. Rep. No. 23 (Univ. South Carolina, Columbia, 2004); in Approximation and Probability (Inst. Math. Pol. Acad. Sci., Warsaw, 2006), Banach Center Publ. 72, pp. 341–366.
V. N. Temlyakov, “Approximation in Learning Theory,” Ind. Math. Inst. Res. Rep. No. 05 (Univ. South Carolina, Columbia, 2005).
V. N. Temlyakov, “Nonlinear Methods of Approximation,” Found. Comput. Math. 3, 33–107 (2003).
V. N. Temlyakov, “Greedy Algorithms in Banach Spaces,” Adv. Comput. Math. 14, 277–292 (2001).
Author information
Authors and Affiliations
Additional information
Published in Russian in Trudy Matematicheskogo Instituta imeni V.A. Steklova, 2006, Vol. 255, pp. 256–272.
Rights and permissions
About this article
Cite this article
Temlyakov, V.N. On universal estimators in learning theory. Proc. Steklov Inst. Math. 255, 244–259 (2006). https://doi.org/10.1134/S0081543806040201
Received:
Issue Date:
DOI: https://doi.org/10.1134/S0081543806040201