Abstract: These days, many places struggle with air pollution, putting the health of young and old at risk for respiratory issues. Forecasting fine-grained air quality in the future is crucial for informing public policy and helping individuals make decisions. Using historical data on air quality, meteorological expertise, and forecasting data, we predict the average air quality for a town for the next seven days, as well as the air quality for each tracking station for the next 48 hours. Our proposal is a deep neural network method called Deep Air, which is based on domain knowledge about air pollutants. We employ a deep cascaded fusion community for longer-term forecasting and a deep distributed fusion network for station-level long-term prediction, and long-term prospects for the city. The previous community used a neural distributed structure as part of the information transformation preprocessing in order to combine diverse city facts and simultaneously collect the direct and indirect components affecting air quality. The latter network examines the dynamic effects of historical, current, and projected future data on air quality using a neural cascaded architecture. Our device specifically integrates three additives— a project scheduler, and a prediction model—to boost the system's efficacy and stability. These additives function through a structure of many challenges. Results from experiments demonstrate the advantages of our proposed approach, which is mostly based on datasets from nine Indian towns over a three-month period.
Index Term: Air excellent prediction, Deep Neural Networks.
Cite:
Dr. S Rajesh, P Bazeer Ahamed, M Deepakkumar, T Subramanian, G Raj Karna,"Predicting Air Quality by Particulate Matter Based on Neural Networks", IJARCCE International Journal of Advanced Research in Computer and Communication Engineering, vol. 13, no. 2, 2024, Crossref https://doi.org/10.17148/IJARCCE.2024.13223.
|
DOI:
10.17148/IJARCCE.2024.13223
