Experiences in the Application of a Multivariate Method for Imputation of Meteorological Data

Main Article Content

José Luis Araya-López

Abstract

Different statistical methods and techniques have been proposed for dealing with missing data. This study discusses the application of the principal components approach for filling hourly meteorological data. In order to test the possibilities that this approach offers, preliminary tests were conducted by random removal of real data in time series. Missing data were predicted using a principal-components algorithm. The results show that this method could predict the missing information with an mean absolute error that is around 1ºC in most of the cases. 

Article Details

How to Cite
Araya-López, J. L. (2014). Experiences in the Application of a Multivariate Method for Imputation of Meteorological Data. Tecnología En Marcha Journal, 27(3), pág. 70–79. https://doi.org/10.18845/tm.v27i3.2068
Section
Artículo científico