Forecasts Using Neural Network versus Box-Jenkins Methodology for Ambient Air Quality Monitoring Data

Abstract

This study explores ambient air quality forecasts using the conventional time-series approach and a neural network. Sulfur dioxide and ozone monitoring data collected from two background stations and an industrial station are used. Various learning methods and varied numbers of hidden layer processing units of the neural network model are tested. Results obtained from the time-series and neural network models are discussed and compared on the basis of their performance for 1-step-ahead and 24-step-ahead forecasts. Although both models perform well for 1-step-ahead prediction, some neural network results reveal a slightly better forecast without manually adjusting model parameters, according to the results. For a 24-step-ahead forecast, most neural network results are as good as or superior to those of the time-series model. With the advantages of self-learning, self-adaptation, and parallel processing, the neural network approach is a promising technique for developing an automated short-term ambient air quality forecast system. INTRODUCTIONAir pollutants exert a wide range of impacts on biological, physical, and economic systems. Their effects on human health are of particular concern. The decrease in

Protected indications of origin as differentiation cue of food products: the Spanish dry-cured ham case

“…Su uso se encuentra ampliamente difundido en múltiples ámbitos, así como en sectores de muy diversa naturaleza (e.g., Chatfield, 1989;Kao y Huang, 2000), demostrándose su validez predictiva y su utilidad en la ayuda para la toma de decisiones (Zou y Yang, 2004). Ello es debido, principalmente, a que las predicciones a corto plazo que proporciona el análisis de series temporales sirve como un excelente complemento a la información estadística de la que dispone la empresa, ya que esta información se suele obtener de modo diferido y con cierto retraso, dificultando la toma de decisiones por parte de la empresa hasta que esta información pueda estar disponible (Otero y Trujillo, 1998).…”

Artificial Neural Networks in Vehicular Pollution Modelling

“…In contrast to deterministic models, statistical models are empirical models. In the model identification stage, the resulting model quality frequently relies on individual's experience and knowledge of the time series statistics [61]. In recent past, time series analysis technique, popularly known as Box-Jenkins' models [50,51,52] have widely been used to describe the dispersion of air pollutants on the local scale [17,53,54].…”

Neural network-based study for predicting ground-level ozone concentration in large urban areas, applied to the Sao Paulo metropolitan area

“…The technique has been applied either to correlate ozone concentration with air quality and meteorological variables, or as a prediction tool [3,[11][12][13][14][15][16][17][18][19][20]. The results constitute clear evidence of the ability of NN models to reproduce the behaviour of complex non-linear systems.…”