The Application of Chaos Theory on Covid-19 Daily Time Series Dataset in Malaysia

Abstract

The third wave of Covid-19 in Malaysia was unpredictable and uncertain due to invisible transmission infection. The daily time series dataset of covid-19 has been analysed to identify the existence of chaotic behaviour for the purpose to develop a prediction model based on chaos theory. Two covid-19 time series datasets are being selected which consist of datasets for new positive cases and new death cases in Malaysia. In order to develop a prediction method based on chaos theory, the analysis of chaotic dynamics is running using the Cao method [E2(m)] and phase space plot. Both methods will undergo phase space reconstruction and then the phase of development of the prediction method using chaos theory. To reconstruct the phase space, two parameters are needed which included time delay () that are obtained through the calculation of average mutual information (AMI) and embedding dimension (d) through the Cao method [E1(m)]. The analysis of chaotic existence shows the presence of chaotic behaviour in the time series data for both datasets. Hence, the time series data for both datasets have been estimated excellently using the local average approximation method where the correlation coefficients in all stations obtained more than 90% accuracy. Hence, chaos theory can be used to model the prediction of the Covid-19 dataset. This research may aid local authorities in their early planning for Covid-19 case management.Covid-19;