Smoothing noisy data with spline functions

By Grace Wahba. 1975

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Component selection and smoothing in multivariate nonparametric regression (2006, methodology): “ procedure and the fully iterated procedure have comparable estimation accuracy. 12 Y. LIN AND H. H. ZHANG 6. Choosingthetuningparameter. The generalized cross-validation proposed by Craven and Wahba [4] is one of the most popular methods for choosing smoothing parameters in the smoothing spline met”
Growth and mortality rates of the fan mussel Pinna nobilis in Lake Vouliagmeni (Korinthiakos Gulf, Greece): a generalized additive modelling approach (2007, methodology): “The optimum degree of smoothing was defined by Generalized Cross Validation (GCV) in the case of Eq. 2 and by minimizing the Un-Biased Risk Estimator (UBRE; Craven and Wahba 1979 ) in the case of Eq. 3, increasing the amount that each model effective degree of freedom counts in the GCV or UBRE scor”
Selection of smoothing parameters inB-spline nonparametric regression models using information criteria (2003, methodology): “Generalized cross-validation introduced by Craven and Wahba (1979) replaces s ~ in (3.”
Comparing parameter choice methods for regularization of ill-posed problems (2011, methodology): “By using a certain weighting of the prediction errors, Wahba [31, 51, 140, 142] derived the GCV method, which has the advantage of being invariant under orthogonal transformations of the data.”
Semiparametric estimates of the relation between weather and electricity sales (1986, methodology): “In the empirical work reported later, the GCV criterion was used because of its relative simplicity and because it was shown by Craven and Wahba (1979) to be an effective method for choosing A for ordinary smoothing splines.”
Monte-Carlo Sure: A Black-Box Optimization of Regularization Parameters for General Denoising Algorithms (2008, background): “GCV is based on the “leave-one-out” principle [2]–[5] and is known to yield λ which asymptotically minimizes (under certain hypotheses) the true MSE [25].”