Comparing error estimation measures for polynomial and kriging approximation of noise-free functions |
| |
Authors: | Tushar Goel Raphael T Hafkta Wei Shyy |
| |
Affiliation: | (1) University of Florida, Gainesville, FL 32611, USA;(2) University of Michigan, Ann Arbor, MI 48109, USA;(3) Present address: Livermore Software Technology Corporation, Livermore, CA, USA |
| |
Abstract: | Error estimation measures are useful for assessing uncertainty in surrogate predictions. We use a suite of test problems to
appraise several error estimation measures for polynomial response surfaces and kriging. In addition, we study the performance
of cross-validation error measures that can be used with any surrogate. We use 1,000 experimental designs to obtain the variability
of error estimates with respect to the experimental designs for each problem. We find that the (actual) errors for polynomial
response surfaces are less sensitive to the choice of experimental designs than the kriging errors. This is attributed to
the variability in the maximum likelihood estimates of the kriging parameters. We find that no single error measure outperforms
other measures on all the problems. Computationally expensive integrated local error measures (standard error for polynomials
and mean square error for kriging) estimate the actual root mean square error very well. The distribution-free cross-validation
error characterized the actual errors reasonably well. While the estimated root mean square error for polynomial response
surface is a good estimate of the actual errors, the process variance for kriging is not. We explore a few methods of simultaneously
using multiple error measures and demonstrate that the geometric means of several combinations of error measures improve the
assessment of the actual errors over individual error measures. |
| |
Keywords: | Error measures Surrogate models Ensemble strategy PRESS Prediction variance |
本文献已被 SpringerLink 等数据库收录! |
|