Regression and Artificial Neural Network Model for Forecasting Air Pollution in winter and Solution Approaches

One of the major problems in Tehran is its increasing pollution that in addition to respiratory and pulmonary problems, affect the learning and the memory of the residents of the city, especially children which every year Tehran municipally officials have come up with different ways to tackle with this problems by some methods such as even-odd scheme for car traffic leading to create less CO2 in the city. In this paper, we use 14 different data from 3 various meteorological stations located in the polluted area of the city for presenting our model. The reason for utilizing the three-month data of the winter in also the phenomenon of the weather inversion through this season in which we confront the most polluted season of the year. Using the data, we propose a regression model as well as a neural network to predict the winter air pollution index and compare the results and error rates with the actual values. These models can be used on a platform designed to make more precise and better decision making, taking into account all parameters that are related to the pollution, in order to reduce air pollution as much as possible.