Honey Yield Forecast Using Radial Basis Functions
Honey yields are difficult to predict and have been usually associated with weather conditions. Although some specific meteorological variables have been associated with honey yields, the reported relationships concern a specific geographical region of the globe for a given time frame and cannot be used for different regions, where climate may behave differently. In this study, Radial Basis Function (RBF) interpolation models were used to explore the relationships between weather variables and honey yields. RBF interpolation models can produce excellent interpolants, even for poorly distributed data points, capable of mimicking well unknown responses providing reliable surrogates that can be used either for prediction or to extract relationships between variables. The selection of the predictors is of the utmost importance and an automated forward-backward variable screening procedure was tailored for selecting variables with good predicting ability. Honey forecasts for Andalusia, the first Spanish autonomous community in honey production, were obtained using RBF models considering subsets of variables calculated by the variable screening procedure.
KeywordsHoney yield Weather Radial basis functions Variable screening
This work has been supported by the Fundação para a Ciência e a Tecnologia (FCT) under project grant UID/MULTI/00308/2013.
- 1.Anuário de Estadistica del Ministerio de Agricultura, Alimentación y Medio Ambiente. http://www.mapama.gob.es/
- 2.Crane, E.: The Archaeology of Beekeeping. Cornell University Press, Ithaca (1983)Google Scholar
- 3.FAOSTAT. http://www.fao.org/faostat/
- 4.Holmes, W.: Weather and honey yields. Scott. Beekeep. 75, 190–192 (1988)Google Scholar
- 6.Hurst, G.W.: Honey production and summer temperatures. Meteorol. Mag. 96, 116–120 (1967)Google Scholar
- 7.Hurst, G.W.: Temperatures inhigh summer, and honey production. Meteorol. Mag. 99, 75–82 (1970)Google Scholar
- 8.Instituto de investigación y formación agraria y pesquera. www.juntadeandalucia.es/agriculturaypesca/ifapa/ria/
- 10.MATLAB 2016a: Natick. The MathWorks Inc., Massachusetts (2016)Google Scholar
- 19.Tu, J.: Cross-validated multivariate metamodeling methods for physics-based computer simulations. In: Proceedings of the IMAC-XXI (2003)Google Scholar
- 20.Tu, J., Jones, D.R.: Variable screening in metamodel design by cross-validated moving least squares method. In: Proceedings of the 44th AIAA (2003)Google Scholar