Optimisation of Interpolation Parameters Using Cross-validation
A number of EU and national policy strategies rely on the use of quality digital elevation model (DEM) data. This chapter tests the reliability and applicability of crossvaliadation in an automatic optimisation of interpolation control parameters on an example of the Regularized Spline with Tension (RST) and elevation data. The input elevation points are derived by random sampling with various densities from grid-based digital elevation models with two different spatial resolutions (25m and 100m). The CV is applied on each sampled set of points to identify the ptimum RST parameters. These parameters are subsequently applied in the interpolation of new surfaces, and their deviations from the original DEMs are analysed.
HOFIERKA Jaroslav;
CEBECAUER Tomas;
SURI Marcel;
2008-03-17
Springer
JRC43643
978-3-540-36730-7,
Additional supporting files
| File name | Description | File type | |
|---|
