Daytime ozone concentration prediction using statistical models

Muhammad Yazid Nasir, . and Nurul Adyani Ghazali, . and Muhammad Izwan Zariq Mokhtar, . and Nor Azam Ramli, . and Norhazlina Suhaimi, . and Noor Faizah Fitri Md Yusof, . Daytime ozone concentration prediction using statistical models. pp. 7-11. ISSN 2672-7226

Ground-level ozone (O‚) has a significant effect on human health when the concentration level exceeds Malaysia Ambient Air Quality Guidelines (MAAQG). This research focuses on daytime ground-level O‚ concentrations in Kemaman Terengganu. The aim of this study is to compare the performance of the multiple linear regression model and the feed forward backpropagation neural network model for predicting daytime O‚ concentrations. This study used hourly average monitoring records from 2009 to 2012. Five performance indicators that are normalized absolute error (NAE) root mean squared error (RMSE) index of agreement (IA) prediction accuracy (PA) and coefficient of determination (R ) were used to evaluate the models performances. The feed forward backpropagation neural network model shows better performances with smaller calculated errors (NAE 0.1729 RMSE 6.7906) and high accuracy (IA 0.9427 PA 0.8054 R 0.8022) than the multiple linear regression. The performances of feed forward backpropagation neural network model can be used for O‚ concentration prediction in the future.

Full text available from: Official URL: http://jssm.umt.edu.my/wp-content/uploads/sites/51...

Abstract

Ground-level ozone (O‚) has a significant effect on human health when the concentration level exceeds Malaysia Ambient Air Quality Guidelines (MAAQG). This research focuses on daytime ground-level O‚ concentrations in Kemaman Terengganu. The aim of this study is to compare the performance of the multiple linear regression model and the feed forward backpropagation neural network model for predicting daytime O‚ concentrations. This study used hourly average monitoring records from 2009 to 2012. Five performance indicators that are normalized absolute error (NAE) root mean squared error (RMSE) index of agreement (IA) prediction accuracy (PA) and coefficient of determination (R ) were used to evaluate the models performances. The feed forward backpropagation neural network model shows better performances with smaller calculated errors (NAE 0.1729 RMSE 6.7906) and high accuracy (IA 0.9427 PA 0.8054 R 0.8022) than the multiple linear regression. The performances of feed forward backpropagation neural network model can be used for O‚ concentration prediction in the future.

[error in script]

Item Type:	Article
AGROVOC Term:	Ozone
AGROVOC Term:	Time
AGROVOC Term:	Prediction
AGROVOC Term:	Statistical methods
AGROVOC Term:	Chemical concentration
AGROVOC Term:	Meteorological data
AGROVOC Term:	Environmental monitoring
AGROVOC Term:	Neural networks
AGROVOC Term:	Air pollution
Depositing User:	Mr. AFANDI ABDUL MALEK
Last Modified:	24 Apr 2025 00:54
URI:	http://webagris.upm.edu.my/id/eprint/8222

Actions (login required)

View Item