The Annals of Statistics

Projection estimation in multiple regression with application to functional ANOVA models

Jianhua Z. Huang

Full-text: Open access


A general theory on rates of convergence of the least-squares projection estimate in multiple regression is developed. The theory is applied to the functional ANOVA model, where the multivariate regression function is modeled as a specified sum of a constant term, main effects (functions of one variable) and selected interaction terms (functions of two or more variables). The least-squares projection is onto an approximating space constructed from arbitrary linear spaces of functions and their tensor products respecting the assumed ANOVA structure of the regression function. The linear spaces that serve as building blocks can be any of the ones commonly used in practice: polynomials, trigonometric polynomials, splines, wavelets and finite elements. The rate of convergence result that is obtained reinforces the intuition that low-order ANOVA modeling can achieve dimension reduction and thus overcome the curse of dimensionality. Moreover, the components of the projection estimate in an appropriately defined ANOVA decomposition provide consistent estimates of the corresponding components of the regression function. When the regression function does not satisfy the assumed ANOVA form, the projection estimate converges to its best approximation of that form.

Article information

Ann. Statist. Volume 26, Number 1 (1998), 242-272.

First available: 28 August 2002

Permanent link to this document

Mathematical Reviews number (MathSciNet)

Digital Object Identifier

Zentralblatt MATH identifier

Primary: 62G07: Density estimation
Secondary: 62G20: Asymptotic properties

ANOVA curse of dimensionality finite elements interaction least squares polynomials rate of convergence regression splines tensor product trigonometric polynomials wavelets


Huang, Jianhua Z. Projection estimation in multiple regression with application to functional ANOVA models. The Annals of Statistics 26 (1998), no. 1, 242--272. doi:10.1214/aos/1030563984.

Export citation


  • BREIMAN, L. 1993. Fitting additive models to data. Comput. Statist. Data. Anal. 15 13 46. Z.
  • CHEN, Z. 1991. Interaction spline models and their convergence rates. Ann. Statist. 19 1855 1868. Z.
  • CHUI, C. K. 1988. Multivariate Splines. SIAM, Philadelphia. Z.
  • DAUBECHIES, I. 1994. Two recent results on wavelets: wavelets bases for the interval, and biorthogonal wavelets diagonalizing the derivative operator. Recent Advance in Wavelet Analy sis 237 258. Z.
  • DE BOOR, C. 1976. A bound on the L -norm of L -approximation by splines in terms of a global 2 mesh ratio. Math. Comp. 30 765 771. Z.
  • DE BOOR, C. 1978. A Practical Guide to Splines. Springer, New York. Z.
  • DEVORE, R. A. and LORENTZ, G. G. 1993. Constructive Approximation. Springer-Verlag, Berlin. Z.
  • DEVORE, R. A. and POPOV, V. 1988. Interpolation of Besov spaces. Trans. Amer. Math. Soc. 305 397 414. Z.
  • DONOHO, D. L. and JOHNSTONE, I. M. 1992. Minimax estimation via wavelet shrinkage. Technical Report 402, Dept. Statistics, Stanford Univ. Z.
  • DONOHO, D. L., JOHNSTONE, I. M., KERKy ACHARIAN, G. and PICARD, D. 1995. Wavelet shrinkage: Z. asy mptopia? with discussion. J. Roy. Statist. Soc. Ser. B 57 301 369. Z. Z.
  • FRIEDMAN, J. H. 1991. Multivariate adaptive regression splines with discussion. Ann. Statist. 19 1 141. Z.
  • FRIEDMAN, J. H. and SILVERMAN, B. W. 1989. Flexible parsimonious smoothing and additive Z. modeling with discussion. Technometrics 31 3 39. Z.
  • HANSEN, M. 1994. Extended linear models, multivariate splines, and ANOVA. Ph.D. dissertation, Univ. California, Berkeley. Z.
  • HASTIE, T. J. and TIBSHIRANI, R. J. 1990. Generalized Additive Models. Chapman and Hall, London. Z.
  • HOEFFDING, W. 1963. Probability inequalities for sums of bounded random variables. J. Amer. Statist. Assoc. 58 13 30. Z.
  • HUANG, J. Z. 1996. Functional ANOVA models for generalized regression. Technical Report 458, Dept. Statistics, Univ. California, Berkeley. Z.
  • HUANG, J. Z. and STONE, C. J. 1997. The L rate of convergence for event history regression 2 with time-dependent covariates. Scand. J. Statist. To appear. Z.
  • HUANG, J. Z., STONE, C. J. and TRUONG, Y. K. 1997. Functional ANOVA models for proportional hazards regression. Unpublished manuscript. Z.
  • KOOPERBERG, C., BOSE, S. and STONE, C. J. 1997. Poly chotomous regression. J. Amer. Statist. Assoc. 92 117 127.
  • KOOPERBERG, C., STONE, C. J. and TRUONG, Y. K. 1995. Hazard regression. J. Amer. Statist. Assoc. 90 78 94. Z.
  • MEy ER, Y. 1992. Wavelets and Operators. Cambridge University Press. Z.
  • ODEN, J. T. and CAREY, G. F. 1983. Finite Elements: Mathematical Aspects. Prentice-Hall, Englewood Cliffs, NJ. Z.
  • OSWALD, P. 1994. Multilevel Finite Element Approximation: Theory and Application. Teubner, Stuttgart. Z.
  • POLLARD, D. 1990. Empirical Processes: Theory and Application. IMS, Hay ward, CA. Z.
  • SCHUMAKER, L. L. 1981. Spline Functions: Basic Theory. Wiley, New York. Z.
  • SCHUMAKER, L. L. 1991. Recent progress on multivariate splines. In Mathematics of Finite Z. Elements and Application VII J. Whiteman, ed. 535 562. Academic Press, London. Z.
  • STONE, C. J. 1982. Optimal global rates of convergence for nonparametric regression. Ann. Statist. 10 1348 1360. Z.
  • STONE, C. J. 1985. Additive regression and other nonparametric models. Ann. Statist. 13 689 705. Z.
  • STONE, C. J. 1994. The use of poly nomial splines and their tensor products in multivariate Z. function estimation with discussion. Ann. Statist. 22 118 171. Z.
  • STONE, C. J., HANSEN, M., KOOPERBERG, C. and TRUONG, Y. 1997. Poly nomial splines and their Z. tensor products in extended linear modeling with discussion. Ann. Statist. 25 1371 1470. Z.
  • STONE, C. J. and KOO, C. Y. 1986. Additive splines in statistics. In Proceedings of the Statistical Computing Section 45 48. Amer. Statist. Assoc., Washington, D.C. Z.
  • TAKEMURA, A. 1983. Tensor analysis of ANOVA decomposition. J. Amer. Statist. Assoc. 78 894 900. Z.
  • TIMAN, A. F. 1963. Theory of Approximation of Functions of a Real Variable. MacMillan, New York. Z.
  • WAHBA, G., WANG, Y., GU, C., KLEIN R. and KLEIN, B. 1995. Smoothing spline ANOVA for exponential families, with application to the Wisconsin epidemiological study of diabetic retinopathy. Ann. Statist. 23 1865 1895.