Title: Applications of errors-in-variables models to wind data
Authors: Stefania Fensore - University of Chieti-Pescara (Italy) [presenting]
Marco Di Marzio - University of Chieti-Pescara (Italy)
Agnese Panzera - University of Florence (Italy)
Charles C Taylor - University of Leeds (United Kingdom)
Abstract: Wind direction data are typically subject to biases of several degrees because measurements are affected by many factors. Then, wind directions can be regarded as circular data observed with error. We discuss some kernel-based methods for estimating circular densities when data are observed with error and propose them to tackle the prediction of wind direction distributions as an errors-in-variables problem.