Time Series Analysis and Prediction of Particulate Matter using Deep Learning Method

DOI

10.2991/978-94-6463-740-3_25How to use a DOI?

Keywords

deep learning; LSTM; RNN; air pollution; Time series

Abstract

With the advent of fast paced industrialization and rapid growth and urbanization, the grave issue of air pollution in urban areas, especially in industrial sectors and in developing countries is at an all-time high. Air pollution can have adverse effects on the health and well-being of humans. It can cause serious diseases and chronic illnesses and may even lead to a sharp decline in the quantity and quality of human life if left unchecked. Various countries have tried to build stations that observe air pollution in order to record the amount of particulate matter around major cities and urban centers. While there are various methods used in the forecasting of air pollution, Machine Learning is one of the most popular techniques that is utilized for the prediction of air pollution as well as steps to counter its increase. Specifically, Deep Learning can be utilized in prediction of air pollution as demonstrated in this paper. Time series analysis, as its name suggests, is a method that is used to analyze a sequence of data values and points that have been collected over a certain period. In this technique, analysts keep a record of the data points that are measured at consistent periods over a fixed interval of time instead of recording the values randomly or sporadically. What distinguishes time series analysis from other methods of data analysis is that it has the ability to show us how the variables have changed over the interval of time. Time is an extremely important variable because it showcases exactly how the data has changed and adjusted over the course of the interval as compared to the final result. Time series analysis usually requires a huge amount of data in order for the results to be consistent and reliable. In this paper, Recurrent Neural Networks (RNN) and Long-Short Term Memory (LSTM) is proposed to predict the air quality from data collected by satellites to estimate the approximate range of air particulate matter and pollutants from the years 2005-2018, Headed by the University of Birmingham and University College London. Model performance obtained was R2=92.6% for LSTM and R2=91.4% for the RNN.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Deepak Gaur
AU  - Rishi Kumar
AU  - M. S. Guru Prasad
PY  - 2025
DA  - 2025/06/25
TI  - Time Series Analysis and Prediction of Particulate Matter using Deep Learning Method
BT  - Proceedings of the 6th International Conference on Deep Learning, Artificial Intelligence and Robotics (ICDLAIR 2024)
PB  - Atlantis Press
SP  - 293
EP  - 302
SN  - 1951-6851
UR  - https://doi.org/10.2991/978-94-6463-740-3_25
DO  - 10.2991/978-94-6463-740-3_25
ID  - Gaur2025
ER  -