An improved methodology for filling missing values in spatiotemporal climate data set
- 290 Downloads
In this paper, an improved methodology for the determination of missing values in a spatiotemporal database is presented. This methodology performs denoising projection in order to accurately fill the missing values in the database. The improved methodology is called empirical orthogonal functions (EOF) pruning, and it is based on an original linear projection method called empirical orthogonal functions (EOF). The experiments demonstrate the performance of the improved methodology and present a comparison with the original EOF and with a widely used optimal interpolation method called objective analysis.
KeywordsMissing value problem Empirical orthogonal functions EOF Selection of singular values Tanganyika Lake
Unable to display preview. Download preview PDF.
- 3.Preisendorfer, R.: Principal Component Analysis in Meteorology and Oceanography. Elsevier, Amsterdam (1988)Google Scholar
- 9.Lendasse, A., Wertz, V., Verleysen, M.: Model selection with cross-validations and bootstraps—application to time series prediction with rbfn models. In: LNCS, no. 2714, pp. 573–580. Springer, Berlin (2003)Google Scholar
- 10.Kohavi, R.: A study of cross-validation and bootstrap for accuracy estimation and model selection. In: Proceedings of the 14th International Joint Conference on Artificial Intelligence, vol. 2, Montreal, 20–25 August 1995Google Scholar
- 14.Gandin, L.S.: Objective Analysis of Meteorological Fields, p. 242. Israel Program for Scientific Translations, Jerusalem (1969)Google Scholar