Research and Realization of Air Quality Grade Prediction Based on KNN

Since entering modern society, people have paid more and more attention to air quality in order to better help predict the air quality level. This paper proposes an air quality grade prediction model based on the K-nearest neighbor algorithm. Firstly, the historical measurement data of air quality is crawled from the relevant weather website and saved to the local CSV file; then the data is read, and the scatter diagram is used to visually display the 6 characteristics that affect the air quality level evaluation; then the K nearest neighbor algorithm is selected, and the difference is adjusted. The parameter training model of, and then through the test set verification, the test accuracy rate is 95.10%. Finally, a set of new data is randomly given, and the prediction results are in line with the expected results, which can be extended to predict the air quality level.