Air Quality Analysis and Prediction Based on Big Data: A Case Study of Tianjin

In this study, air quality data from Tianjin spanning the period from January 1, 2018, to April 20, 2024, were analyzed. A Gaussian Naive Bayes model was employed to predict the classification levels of the Air Quality Index(AQI). The results demonstrate that the Gaussian Naive Bayes model possesses robust predictive capabilities. The model predicted that the AQI for the subsequent twenty days would range between 50 and 150, achieving a prediction accuracy of 85%. Furthermore, the analysis revealed significant fluctuations in the concentrations of PM10 and PM2.5 in Tianjin. Currently, PM2.5 is identified as the critical factor influencing air quality classification.