ARTIFICIAL NEURAL NETWORK BASED PREDICTION OF PM2.5 USING NEIGHBORING INFORMATION

Abstract

The negative impact of the fine particulate matter (PM2.5) on health and economic is undeniable. There are many attempts to prediction the concentration of the particle.  The data for the prediction model comes from only the area of observation. However, the particle in a certain area  can originate from the observed area or come from the nearby area. In this article, the artificial neural network based prediction method using the metrological data from the neighbor area is proposed. The proposed method combines metrological data from local area and adjacent location and uses the artificial neural network as a prediction model. Then, the model is challenged against the PM2.5 in Chiang Mai area. The result indicates that the proposed method performs better than the tradition machine learning models: linear regression, support vector machine, and multi-layer perceptron.