The purpose of this project was to study air pollution in Beijing (China) using various methods learned from the data mining course. Trying to find which pollutants most contributed to PM2.5 pollution. PM2.5 pollution is particulate matter smaller than 2.5 microns, which can lead to several negative health effects，hoping to find the causes of serious air pollution in Beijing. My data was getting form the government official website and you can find it with the link (http://beijingair.sinaapp.com/)

In this project, I use C50, KNN, Naive Bayes and Random Forest to test the data, for the four selected algorithms after cacluated, I found random forest is the most accurate one. From which we were able to find that NO2 and CO were the worst, and thus a major cause of the air pollution.

If you want to try my test data, download the final4.csv dataset and the analysis_code.R change the file addrees to your local dataset and you will get the calculat result.

All the suggestions and questions are weclome, thank you