Air quality index prediction with influence of meteorological parameters using machine learning model for IoT application

The growing of industries and urban areas has induced air pollution that in turn led to several impacts in humans and environment over these years globally. Vast proportion of fine particulate matter (PM) along with its variants, namely, PM2.5, PM 10 and gasoline pollutants (i.e. NOX, SO2, CO) is associated with lung cancer, cardiovascular disease, respiratory and metabolic disorders. Mainly, the air pollution is determined by both PM and gasoline pollutants. However, the meteorological parameters like temperature, humidity and wind have influenced the PM and gasoline pollutants. Prediction of air quality index (AQI) based on the metrological parameters is a complex task. Although many attempts are made to predict the AQI, the meteorological influenced forecast have not been probed further. In this work, a multivariant regressive function is developed as multiple linear regressive (MLR) model to predict AQI based on the correlation of two time series–dependent variables for air pollutants and meteorological parameters. The MLR model gives higher efficacy in AQI prediction while comparing the existing algorithms. Predicted AQI is hosted in social network for welfare of the nation and to initiate an awareness in people about the degradation of air quality and its associated health issues.