A model to forecast the risk periods of Plantago pollen allergy by using the ANN methodology

Abstract

Some biological particles present in the atmosphere, such as pollen grains, give rise to human health problems, allergies, and infections. In view of the recognized special allergenic ability of Plantago pollen grains, a model based on an artificial neural network (ANN) was developed in this work in order to forecast the Plantago airborne pollen concentration. The proposed model uses data from Plantago pollen and the main meteorological variables recorded during 16 years (1993–2008) in the city of Ourense (northwest Spain). Its accuracy was tested during the years 2009 and 2010 with a prediction horizon of 2 days in advance. The model was applied in the atmosphere of the city of Ourense (Spain). Obtained results show that ANN model provides good results against other classical mathematical methodologies, which do not convergence so well. The forecasted pollen concentrations here are applied to allergology because they allow taking into account preventive measures in risk pollinosis suffers population.