S. J. Yakowitz and F. Szidarovszky. A Comparison of Kriging with Nonparametric Regression Methods. J. Multivariate Analysis, 16:21--53, 1985. 36  Home/Search Document Not in Database Summary Related Articles Check |
This paper is cited in the following contexts:
Bayesian Classification with Gaussian Processes - Williams, Barber (1998) (23 citations) (Correct)
.... if we consider the infill asymptotics (see, e.g. 3] where the number of data points in a bounded region increases, then a local average of the training data at any point x will provide a tightly localized estimate for (x) and hence y(x) this reasoning parallels more formal arguments found in [29]) Thus we would expect the distribution P (y) to become more Gaussian with increasing data. In detail, we used the basis functions erf(x) for = 0:41; 0:4; 0:37; 0:44; 0:39] These were used to interpolate oe(x) at x = 0; 0:6; 2; 3:5; 4:5; 1] 27 Appendix B: Derivatives of log P a ....
S. J. Yakowitz and F. Szidarovszky. A Comparison of Kriging with Nonparametric Regression Methods. J. Multivariate Analysis, 16:21--53, 1985. 36
On The Effect Of Covariance Function Estimation On The.. - Putter, Young (Correct)
....Although an extremely important practical issue, the effect of estimating C is still not all that well understood. There are essentially two approaches to assessing the influence of misspecifying or approximating the covariance function. The first of these (Diamond Armstrong 1984, Warnes 1986, Yakowitz Szidarovsky 1985) is effectively a numerical analysis. Since the kriging weights are determined by solving linear equations involving the covariance matrix of Z(x 1 ) Z(x n ) approximating the covariance function results Accuracy of kriging predictors with estimated covariance 3 in a perturbation of ....
....the number of observations changes. What s worse, typically C and an estimated covariance function C n will not be equivalent for any finite n. Thus, the question remains open how approximating the true covariance function C by a sequence C n affects the accuracy of the kriging predictor. Since Yakowitz Szidarovsky (1985, p. 39) uttered the remark that we regard the situation as a (perhaps unfillable) lacuna in kriging theory , to the best of our knowledge the problem has not been solved to a satisfactory degree. 4 H. Putter and G. A. Young The aim of this note is to study the effect of estimating the ....
Yakowitz, S. J. & Szidarovsky, F. (1985), `A comparison of kriging with nonparametric regression methods', J. Multivar. Anal. 16, 21--53.
Cokriging, Kernels, And The SVD: Toward Better Geostatistical.. - Long (1994) (Correct)
....in the one dimensional case, using 25 scattered data locations on the interval [0,1] Estimation at x= 4 with four different sets of locations. significant improvement in the use of these techniques. Kriging and kernels have been seen as estimation opponents in some work: Yakowitz and Szidarovszky [99] compared kriging (with correct and misspecified variogram models) with a gaussian smoothing kernel with bandwidth corrections. Using several test data sets, they concluded that kriging gave better estimates when the variogram model was correctly specified, but that the kernel did a better job ....
S. J. Yakowitz and F. Szidarovszky. A comparison of kriging with nonparametric regression methods. Journal of Multivariate Analysis, 16:21--53, 1985.
Bayesian Classification with Gaussian Processes - Williams, Barber (1998) (23 citations) (Correct)
.... we consider the infill asymptotics (see, e.g. 3] where the number of data points in a bounded region increases, then a local average of the training data at any point x will provide a tightly localized estimate for (x) and hence y(x) this reasoning parallels more formal arguments found in [29]) Thus we would expect the distribution P (y) to become more Gaussian with increasing data. Appendix B: Derivatives of log P a (tj ) wrt . For both the HMC and MPL methods we require the derivative of l a = log P a (tj ) with respect to components of , for example k . This derivative will ....
S. J. Yakowitz and F. Szidarovszky. A Comparison of Kriging with Nonparametric Regression Methods. J. Multivariate Analysis, 16:21--53, 1985.
Convergence of a Stochastic Rootfinding Procedure - Burton Simon Department (Correct)
No context found.
Yakowitz, S.J. and Szidarovszky, F., (1985), A Comparison of Kriging with Nonparametric Regression Methods, Journal of Multivariate Analysis, Vol. 16, 21-53. 19 Robbins-Monro oe = :1 oe = 1 oe = 10 slope se(50) se(100) se(200) se(50) se(100) se(200) se(50) se(100) se(200)
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC