The Analysis Of Urban Air Quality And Economy Condition Based On Decision Tree Algorithm

In this paper we will introduce the basic classification theories of data mining, andstudy the indicators data of air quality and economic development of31cities from2003to2011. Based on decision tree classification algorithm, we select a few basic factorswhich affect air quality to make a classification on air pollution problems and derive thedecision tree classification model which is able to predict the classification of the airquality. Furthermore, we know that inhalable particulate matter、sulfur dioxide andindustrial dust emissions are the main factors which affect the air quality.According to the result of the classification, we select three main factors and fiveeconomic index (per capita GDP, proportion of primary industry of GDP, proportion ofsecondary industry of GDP, proportion of tertiary industry of GDP, number of industrialenterprises) to make canonical correlation analysis between air quality and economicconditions of31cities, and obtain canonical correlation factors of air quality andeconomic conditions about the two kinds of cities with the aid of SAS. By comparison,we can conclude that there exists difference in the correlation coefficients of sulfurdioxide and economic status indicators between the cities, especially the proportionaldistribution of the three major industries. In addition, the main contribution factors of thefirst canonical variables are also different. The results could provide us reference aboutair pollution and economic development issues of these cities.