MORTALITY ANALYSIS IN RELATION WITH WEATHER AND AIR POLLUTION USING THE COMPOUND NEURAL NETWORK MODEL

摘要

Despite extensive research of the meteorological effectsrnon air pollution trends and pollutant effects on publicrnhealth, there has been no attempt to combine bothrnmeteorological factors and pollutant factors to estimaternthe levels of risk to human health. This is not simply arnmatter of the number of the predictors but also therncomplexity of their interdependencies. To model thisrncomplex multivariate, double layered relationship, arncompound neural network has been developed to modelrnmultivariate relationships. Experiments have beenrnconducted, performance and stability have been comparedrnto conventional Multi-Layer Perceptron Neural Networksrn(MLPNN). The results are discussed.