Abstract
The computation ofL 1 smoothing splines on large data sets is often desirable, but computationally infeasible. A locally weighted, LAD smoothing spline based smoother is suggested, and preliminary results will be discussed. Specifically, one can seek smoothing splines in the spacesW m (D), with [0, 1]n⊆D. We assume data of the formy i =f(t i )+ε i ,i=1,..., N with {t i } N i=1 ⊂D, the ε i are errors withE(ε i )=0, andf is assumed to be inW m . An LAD smoothing spline is the solution,s λ, of the following optimization problem
whereJ m (g) is the seminorm consisting of the standard sum of the squaredL 2 norms of themth partial derivatives ofg. Such an LAD smoothing spline,s λ, would be expected to give robust smoothed estimates off in situations where the ε i are from a distribution with heavy tails. For fixed λ>0, the solution to such a problem is known to be a thin plate spline onW m , and hences λ is assumed to be of the form\(s_\lambda = \sum\nolimits_{\nu = 1}^M {d_\nu } \phi _\nu + \sum\nolimits_{i = 1}^N {c_i } \zeta _i \) where\(\zeta _i (t) = R_1 (t_i ,t),R(s,t) = R_0 (s,t) + R_1 (s,t)\) is the reproducing kernel forW m (D), R 1 (t i ,t)=projW 0 m R(t i ,t), and the functions {φ v } M v=1 span the Kern (proj W 0 m )=Kern(J m ). Optimality conditions definings λ as the solution to (1) yield an algorithm for its computation. However, this computation becomes unwieldy whenN≃O(103). A possible remedy is to solve “local” problems of the form of (1), on neighborhoods of “size”b, and to blend these locally optimal LAD splines together producing a globally smooth estimator. Two smoothing parameters (the global value of “λ”, and the “local neighborhood” size “b”) should preferably have data driven, cross validated, choice.
Similar content being viewed by others
References
P.M. Anselone and P.J. Laurent, A general method for the construction of interpolating or smoothing spline-functions, Numer. Math. 12 (1968) 66–82.
P. Bloomfield and W.L. Steiger,Least Absolute Deviations: Theory, Applications, and Algorithms, volume 6 ofProgress in Probability and Statistics (Birkhauser, Boston, 1983).
K.W. Bosworth and U. Lall, Anl 1 smoothing spline algorithm with cross validation, Numer. Algor. 5 (1993) 407–417.
L. Breiman, The Π method for estimating multivariate functions from noisy data, Technometrics 33 (1991) 125–160.
A. Brooke, D. Kendrick and A. Meeraus,GAMS: a User's Guide (Scientific Press, 1988).
A. Buja, T. Hastie and R. Tibshirani, Linear smoothers and additive models, Ann. Statist. 17 (1989) 453–555.
W.S. Cleveland, S.J. Devlin and E. Grosse, Regression by local fitting, J. Econometrics 37 (1988) 87–114.
N. Cressie, Kriging nonstationary data, J. Amer. Statist. Assoc.: Applications 81(395) (1986) 625–634.
F. de Hoog and M. Hutchinson, An efficient method for calculating smoothing splines using orthogonal transformations, Numer. Math. 50 (1987) 311–319.
J. Dongarra, J. Bunch, C. Moler and G. Stewart,Linpack Users' Guide, Society for Industrial and Applied Mathematics (1979).
J. Friedman, Multivariate adaptive regression splines, Ann. Statist. 19 (1991) 1–141.
P.E. Gill, W. Murray, M.A. Saunders and M.H. Wright, User's guide for npsol (version 4.0). Technical report, Department of Operations Research, Stanford University SOL 86-2 (1986).
M. Hutchinson and F. de Hoog. Smoothing noisy data with spline functions, Numer. Math. 47 (1985) 96–106.
C. Jennen-Steinmetz and T. Gasser, A unifying approach to nonparametric regression estimation, J. Amer. Statist. Assoc.: Theory and Methods 83(404) (1988) 1084–1089.
J. Keener,Principles of Applied Mathematics (Addison-Wesley, 1988).
C.A. Michelli, Interpolation of scattered data: Distance matrices and conditionally positive definite matrices, In S. P. Singh et al. (eds.),Approximation Theory and Spline Functions (D. Reidel, 1984) pp. 143–145.
H.-G. Müller and U. Stadtmüller. Variable bandwidth kernel estimators of regression curves, Ann. Statist. 15(1) (1987) 182–201.
H.-G. Müller, Weighted regression and kernel methods for nonparametric curve fitting, J. Amer. Statist. Assoc. 82(397) (1987) 231–238.
G. Nielson, Multivariate smoothing and interpolating splines, SIAM J. Numer. Anal. 11(2) (1974) 435–446.
E. Quak and N. Weyrich, Decomposition and reconstruction algorithms for spline wavelets on a bounded interval, Technical report, Center for Approximation Theory, Texas A. & M. University, CAT Report 294 (1993).
R.T. Rockafellar,Convex Analysis (Princeton, 1970).
G. Wahba,Spline Models for Observational Data (SIAM, 1990).
G.A. Watson,Approximation Theory and Numerical Methods (Wiley, 1980).
S. J. Yakowitz and F. Szidarovsky, A comparison of kriging with nonparametric regression methods, J. Multivar. Anal. 16(1) (1985) 21–53.
J.L. Zhou and A. Tits, User's guide for fsqp version 3.0: A FORTRAN code for solving constrained nonlinear (minimax) optimization problems, generating iterates satisfying all inequality and linear constraints, Technical Report TR-90-60rlf, Univ. of MD. S.R.C. (1990)
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Bosworth, K.W., Lall, U. LOWLAD: a locally weightedL 1 smoothing spline algorithm with cross validated choice of smoothing parameters. Numer Algor 9, 85–106 (1995). https://doi.org/10.1007/BF02143928
Issue Date:
DOI: https://doi.org/10.1007/BF02143928