Air Quality Forecast using Convolutional Neural Network for Sustainable Development in Urban Environments

Abstract

The expansion of IT based technology has widely gathered and enhanced all the eras of applicable domain to maintain the sustainable transformation. In fact, the new evolving technology has broadened its scope for sustainable environment exploiting diverse technological advancement utilizing future prediction modelling. Moreover, smart cities are grounded on the concept of prioritizing development goal discussed among all the nations, the significance sustainable areas are subjected on air, water and energy. Hence, current study of approach relies specifically on atmospheric pollutants which arise due to industrial or high-powered transport emission. The aim of the study, was to utilize the CNN model on the air quality dataset to detect patterns for future prediction modelling. The proposed study, is implemented on two phases where the first phase focuses on preprocessing and data analysis, whereas the Second phase is used for the purpose of testing the model accuracy where the data has been classified on accurate models. The overall model is implemented using Python programming and scripting language. Further, the inclusive result analysis were implemented and gathered to discover the highest level of CO, SO2 and NO2 levels in past five years among the different cities of India from 2015-2020. The cities with highest level of air pollutants and PM2.5 and PM10 effects on the overall Indian cities. Nonetheless, the accuracy of the model is measured to determine the applicability of algorithm.