Prediction of PM10 Time Series at Industrial Area Through an Improved Local Linear Approximation Method

Predicting Haze Phenomenon using Chaos Theory in Industrial Area in Malaysia

“…The haze prediction can be employed using various methodologies, namely chaos (Hamid 2020;Hamid & Noorani 2014), multiple linear regression (Abdullah et al 2017;Chen & Wang 2019), neural network (Pinghua 2018;Zhang et al 2021), long-short term memories (Liu et al 2019;Wu et al 2021), and deep learning (Hasmarullzakim & Abdullah 2018;Idris & Yassin 2021). From study of Shahizam et al (2021), Hamid (2020), Hamid and Noorani (2017) and several other papers showed that prediction using chaos method produced small error and high correlation coefficient. Therefore, since the chaos approach can be used to produce accurate predictions and only PM10 time series is required in prediction, this method will be used in this study.…”

“…There are several methods to determine if the dynamics of a time series is chaos, such as Cao method and phase space plot (Hamid 2018;Adenan et al 2021;Zhao et al 2023), Lyapunov exponents (Hu et al 2023;Pino-Vallejo et al 2018;Wang et al 2023), and correlation dimension (Saeed et al 2017;Yu et al 2013). Study by Shahizam et al (2021), Hamid (2020), Hamid and Noorani (2017) used Cao method and phase space plot to do in chaos method and they had produced good prediction model. Thus, based on these previous successful outcomes of Cao method and phase space plot, the same methodology will be employed in this study.…”