Technical note: Estimating unbiased transfer-function performances in spatially structured environments
Conventional cross validation schemes for assessing transfer-function performance assume that observations are independent. In spatially structured environments this assumption is violated, resulting in over-optimistic estimates of transfer-function performance. <i>H</i>-block cross vali...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Copernicus Publications
2016-05-01
|
| Series: | Climate of the Past |
| Online Access: | http://www.clim-past.net/12/1215/2016/cp-12-1215-2016.pdf |
| Summary: | Conventional cross validation schemes for assessing transfer-function
performance assume that observations are independent. In spatially
structured environments this assumption is violated, resulting in
over-optimistic estimates of transfer-function performance. <i>H</i>-block
cross validation, where all samples within <i>h</i> kilometres of the test samples are
omitted, is a method for obtaining unbiased transfer-function performance
estimates. In this study, we assess three methods for determining the
optimal <i>h</i>. Using simulated data, we find that all three methods result in
comparable values of <i>h</i>. Applying the three methods to published transfer
functions, we find they yield similar values for <i>h</i>. Some transfer functions
perform notably worse when <i>h</i>-block cross validation is used. |
|---|---|
| ISSN: | 1814-9324 1814-9332 |