Citation

Ratchagit, Manlika Forecasting PM2.5 Concentrations in Chiang Mai using Machine Learning Models. International Journal on Robotics, Automation and Sciences, 6 (2). ISSN 2682-860X

Text 1042-Article Text-10431-6-10-20250519.pdf - Published Version Restricted to Repository staff only Download (496kB)

Abstract

Particulate matter 2.5 poses a significant threat to human life. Over the past decade, there has been a significant increase in the number of articles dedicated to studying and forecasting PM2.5 concentrations. Thailand, particularly Chiang Mai, has elevated levels of dangerous PM2.5 throughout the hot season. The primary objective of this study is to evaluate the efficacy of three widely used machine learning models, namely artificial neural network (ANN), long short-term memory network (LSTM), and convolutional neural network (CNN), in predicting the levels of PM2.5 particles in Chiang Mai. The raw data are obtained from the Pollution Control Department, Ministry of Natural Resources and Environment Thailand between January 2014 and June 2023, a total of 3,468 observations. We split the data into three sets namely, training, validation, and test sets. The criterion to evaluate three machine learning techniques is the median absolute error. The experimental results confirm that all three machine learning models provide similar movements of PM2.5 dust pollution. Moreover, the artificial neural network technique provides better results than the others regarding error measurement.

Item Type:	Article
Uncontrolled Keywords:	PM 2.5 Concentrations, Machine Learning, Artificial Neural Network, Long Short-Term Memory Network, Convolutional Neural Network
Subjects:	Q Science > QA Mathematics > QA71-90 Instruments and machines
Divisions:	Others
Depositing User:	Ms Suzilawati Abu Samah
Date Deposited:	11 Jul 2025 03:36
Last Modified:	11 Jul 2025 03:36
URII:	http://shdl.mmu.edu.my/id/eprint/14270

Downloads

Edit (login required)