The Ordinal Classification And Regression Model And Its Application In Judgenment The Rank Of Air Quality

This paper made a preliminary discussion on the structure and the applicability conditions of the ordinal classification and regression model. We applied the model to analyze the influential factors for the air quality of Xi’an City in2014. Taking advantage of the idea and method of ordinal classification and regression model, we removed the non-significant variables, and derived the ordinal classification and regression model of the air quality level of Xi’an City with respect to SO2, PM10, O38hour and PM2.5. Then, we concluded that the air quality level of Xi’an City was significantly related with SO2, PM10, O38hour, PM2.5, especially with the lowest temperature of Xi’an City.The significance of this study lies in the following aspects:(1) The use of raw data related to air quality level, a thorough understanding of the nature of ordinal regression.(2) Use ordinal regression model results given Xi’an air quality model provides a more simple and easy calculation.(3) In fact, a lot of criteria pollutants is we can not find, we can use these "incomplete" indicators to predict air quality.(4) Parallel to air quality forecasts, we can change the situation in accordance with local pollutants and future weather conditions, predict future air quality situation, timely and appropriately reduce or reduce the emission of pollutants, so as to achieve the monitoring of air pollution from passive to active respond to control air pollutants.We found that we could describe the air quality alteration of Xi’an City just with fewer variables, which has certain reference meaning and practical value.