Keywords
feedforward neural networks, pm10, time series forecast
Start Date
1-7-2004 12:00 AM
Abstract
PM10 constitutes a major concern for Milan air quality. We presents a series of results obtainedapplying different neural networks approaches to the PM10 prediction problem. The 1-day ahead predictionshows a satisfactory level of accuracy, which may be further improved if a proper deseasonalization approach isadopted, thus transferring some a priori knowledge in the data pre-processing step. Then, we tackle the problemof the 2-days ahead prediction; in order to optimize the neural network architecture identification procedure,we try a pruning approach besides the usual trial and error one. Prediction performances are very close betweenthe two models, and denote a significant decrease of accuracy with respect to the 1-day case, even though somemeteorological improper (i.e. future measures) input is added to the model structure.
Artificial neural networks prediction of PM10 in the Milan area
PM10 constitutes a major concern for Milan air quality. We presents a series of results obtainedapplying different neural networks approaches to the PM10 prediction problem. The 1-day ahead predictionshows a satisfactory level of accuracy, which may be further improved if a proper deseasonalization approach isadopted, thus transferring some a priori knowledge in the data pre-processing step. Then, we tackle the problemof the 2-days ahead prediction; in order to optimize the neural network architecture identification procedure,we try a pruning approach besides the usual trial and error one. Prediction performances are very close betweenthe two models, and denote a significant decrease of accuracy with respect to the 1-day case, even though somemeteorological improper (i.e. future measures) input is added to the model structure.
