Study of the trend pattern of COVID-19 using spline-based time series model: a Bayesian paradigm

Kumar, J.; Agiwal, Varun; Yau, Chun Yip

doi:10.1007/s42081-021-00127-x

“…As a result, these models struggle to accurately identify key inflection points, exhibit significant lags, and face challenges when it comes to generalization. To enhance the dependability of their predictions, Kumar et al [22] proposed the use of a spline function to segment the non-linear epidemic time series into different growth stages and predict it at different stages of spread of the infection with a linear modeling approach, which reduces the difficulty of prediction. Some researchers synergistically amalgamated the strengths of several models to devise hybrid models [23][24][25][26][27].…”

Section: Literature Reviewmentioning

confidence: 99%

Ensemble Prediction Method Based on Decomposition–Reconstitution–Integration for COVID-19 Outbreak Prediction

Ke,

Lu

2024

Mathematics

2

0

View full text Add to dashboard Cite

Due to the non-linear and non-stationary nature of daily new 2019 coronavirus disease (COVID-19) case time series, existing prediction methods struggle to accurately forecast the number of daily new cases. To address this problem, a hybrid prediction framework is proposed in this study, which combines ensemble empirical mode decomposition (EEMD), fuzzy entropy (FE) reconstruction, and a CNN-LSTM-ATT hybrid network model. This new framework, named EEMD-FE-CNN-LSTM-ATT, is applied to predict the number of daily new COVID-19 cases. This study focuses on the daily new case dataset from the United States as the research subject to validate the feasibility of the proposed prediction framework. The results show that EEMD-FE-CNN-LSTM-ATT outperforms other baseline models in all evaluation metrics, demonstrating its efficacy in handling the non-linear and non-stationary epidemic time series. Furthermore, the generalizability of the proposed hybrid framework is validated on datasets from France and Russia. The proposed hybrid framework offers a new approach for predicting the COVID-19 pandemic, providing important technical support for future infectious disease forecasting.

show abstract

“…Additionally, many modelling studies were done to understand the nature of the pandemic's spread and predict future behaviour. These modelling studies are statistical models such as autoregressive (AR) and Autoregressive Integrated Moving Average (ARIMA), analytical models such as the Susceptible Infected Removed (SIR) model, the Susceptible Exposed Infected Recovered (SEIR) model, and numerical models that use machine learning (ML) and arti cial intelligence (AI) methods [2][3][4][5][6][7][8][9][10]. The parameter values used in these models are determined by comparing data obtained from the World Health Organization or COVID-19 research centres with case data.…”

Section: Introductionmentioning

confidence: 99%

An explication of the 800-day COVID-19 pandemic spread behaviour of seven countries from different continents and the world total in a non-linear time series framework

Tatlipinar

¹

,

KAVAL

²

2023

Preprint

0

View full text Add to dashboard Cite

The time series consisting of 800-day COVID-19 spread data from USA, Germany, Brazil, India, Japan, Egypt, Turkey, and world total taken from the Our World in Data database, was analysed within the framework of non-linear time series. Correlogram diagrams, Fourier power spectra and Lyapunov exponents were examined for each series and it was seen that they did not behave linearly. For these non-linear time series, the lag time and embedded dimension were calculated and 3-dimensional phase spaces for each case were constructed. By examining the constructed phase space profiles, the spread dynamics of COVID-19 in each country and the world total is discussed comparatively. As a result of the phase space analysis, it was seen that the spread of COVID-19 was complex and three different complex behaviour patterns emerged according to the examined countries. This behavioral decomposition was also seen in the correlogram diagrams of the countries, the Fourier power spectrum and the Lyapunov exponents. The nonlinear time series method we used will contribute to the understanding of the qualitative characteristics of the complex behaviour of the COVID-19 pandemic.

show abstract

“…Additionally, many modelling studies were done to understand the nature of the pandemic's spread and predict future behaviour. These modelling studies are statistical models such as autoregressive (AR) and Autoregressive Integrated Moving Average (ARIMA), analytical models such as the Susceptible Infected Removed (SIR) model, the Susceptible Exposed Infected Recovered (SEIR) model, and numerical models that use machine learning (ML) and artificial intelligence (AI) methods [2][3][4][5][6][7][8][9][10]. The parameter values used in these models are determined by comparing data obtained from the World Health Organization or COVID-19 research centres with case data.…”

Section: Introductionmentioning

confidence: 99%

An explication of the 800-day COVID-19 pandemic spread behaviour of seven countries from different continents and the world total in a non-linear time series framework

KAVAL

¹

,

Tatlipinar

²

2023

Preprint

0

View full text Add to dashboard Cite

The time series consisting of 800-day COVID-19 spread data from USA, Germany, Brazil, India, Japan, Egypt, Turkey, and world total taken from the Our World in Data database, was analysed within the framework of non-linear time series. Correlogram diagrams, Fourier power spectra and Lyapunov exponents were examined for each series and it was seen that they did not behave linearly. For these non-linear time series, the lag time and embedded dimension were calculated and 3-dimensional phase spaces for each case were constructed. By examining the constructed phase space profiles, the spread dynamics of COVID-19 in each country and the world total is discussed comparatively. As a result of the phase space analysis, it was seen that the spread of COVID-19 was complex and three different complex behaviour patterns emerged according to the examined countries. This behavioral decomposition was also seen in the correlogram diagrams of the countries, the Fourier power spectrum and the Lyapunov exponents. The nonlinear time series method we used will contribute to the understanding of the qualitative characteristics of the complex behaviour of the COVID-19 pandemic.

show abstract

Study of the trend pattern of COVID-19 using spline-based time series model: a Bayesian paradigm

Cited by 9 publications

References 25 publications

Ensemble Prediction Method Based on Decomposition–Reconstitution–Integration for COVID-19 Outbreak Prediction

Ensemble Prediction Method Based on Decomposition–Reconstitution–Integration for COVID-19 Outbreak Prediction

An explication of the 800-day COVID-19 pandemic spread behaviour of seven countries from different continents and the world total in a non-linear time series framework

An explication of the 800-day COVID-19 pandemic spread behaviour of seven countries from different continents and the world total in a non-linear time series framework

Contact Info

Product

Resources

About