Air Pollution Hotspot Identification and Pollution Level Prediction in the City of Delhi

Abstract

In this paper, we use various methods and algorithms to detect air pollution hotspots and predict pollution levels in a selected area in the city of Delhi. Time series AQI data is collected through the CPCB sensors in Delhi. Classification of hotspots is done using SVM, and the time series analysis based on pollutants like PM2.5, PM10, CO, NO data samples is done using LSTM and PROPHET. Pollution levels of a day in the future are predicted using the said models.