A comparative study on air quality index by machine learning algorithms

Our atmosphere is predominantly consists of two important gases that are important for livelihood, namely Oxygen and Nitrogen. Clean air is essential to stay healthy as well as to maintain a good environment. Air Quality is measured using a metric called Air Quality Index (AQI). It is a number that tells which pollutant is present in the air, its percentage and how it affects our health and Ecosystem. It keeps a tab on major Air Pollutants namely PM10,PM2.5,No2,So2,CO,O3,NH3,Lead. Besides these, Emission of Firecrackers also made an impact of it by releasing harmful chemicals and organic pollutants. In the present study, we had used machine learning models like Linear Regression,Support Vector Machine,Artificial Neural Networks, Decision Trees and Random Forest. The results showed that ANN outperformed other models with the metrics R2 (1),MAE (1), MAPE (0.7142),RMSE(1) for the city Gummidipoondi. This helps government to take preventive steps that worsens with all reasons related to weather conditions,vehicle emissions and local pollutants that create public health emergencies related to many respiratory and cardiovascular diseases.