Linear Regression and Artificial Neural Network (ANN)-based Approaches to Predict Air Pollution

Sharnil Pandya*, Hemant Ghyvat*, Ketan Kotecha*, Prosanta Gope*

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingChapterpeer-review

2 Citations (Scopus)

Abstract

Impacts of Air pollution on the atmosphere and the Health of humans have been a looming issue of the 21st century. Recently, previous studies have conducted a few types of research in the fields of air pollution. Still, the fields of air pollution monitoring and Prediction are considered as open research problems. In the undertaken study, we have presented a novel Pollution Weather System (PWS), which can measure various pollutants such as PM2.5, PM10, CO, O3, and NO2. For the conducted experiments, a real-time air pollution dataset has been used. The investigation results validate the success of the proposed PWS system. We have presented the PWS air pollution, prediction model. In the conducted experiments, linear Regression and ANN-based AQI prediction have been performed. The presented study also found that the customized version of the linear regression methodology is more suitable for air prediction-related applications than the customized version of the ANN algorithm used in the conducted experiments. In the end, a calculation of the Air Quality Index (AQI) has been represented.

Original languageEnglish
Title of host publicationEncyclopedia of Sensors and Biosensors
EditorsMehmet R. Yuce
Place of PublicationAmsterdam, Netherlands
PublisherElsevier
Pages497-511
Number of pages15
Volume3
Edition1st
ISBN (Electronic)9780128225486
ISBN (Print)9780128225493
DOIs
Publication statusPublished - 1 Jan 2023
Externally publishedYes

Cite this