A REVIEW ON DYNAMIC FORECASTING OF AIR POLLUTION IN DELHI ZONE

The high level of pollutants in the surrounding air in 2016-2017 deteriorated the air quality in Delhi at an alarming rate. Future air quality was predicted by analyzing our previous study and we analyzed the data. Forecasting urban air pollution becomes an essential alternative to reduce its harmful consequences. Several machine learning technologies have been adopted Air quality forecasts. In this document, we implement various classification and regression techniques in linear form Regression, ODD regression, random forest regression, Decision tree regression, vector regression support, Artificial neural networks and pulse, gradient regressionAdaptive pulse regression for air quality index prediction Among the main pollutants are PM2.5, PM10, CO, NO2, SO2 and O3. The techniques are then evaluated using the RMS error, mean absolute error and R2, indicating the support vector regression and artificial neural networks are best suited expect New Delhi air quality. The air quality in the Indian capital Delhi has been severe in recent years. A big number people diagnosed with asthma and other breathing problems. The main reason behind this the high concentration of lethal PM2.5 particles dissolved in the atmosphere. Good model predicting the level of concentration of these dissolved particles can help better prepare the population for prevention and safety strategies to save them from many health related diseases. This work aims to predict PM2.5 concentration levels in different areas of Delhi by hour, with time series analysis applied slope, based on various atmospheric and surface factors, such as wind speed and atmospheric temperature,Pressure, etc. Analysis data is obtained from various weather monitoring sites previously installed in the city Indian Meteorological Department (IMD). A regression model has been proposed which uses an additional tree regression AdaBoost, to promote more. Pilot the comparative study with the most recent work and results indicates the efficiency of the proposed model.