COVID-19 prediction using time series modelling in patients in a rural health clinic

Authors

  • María Fernanda Cueva Moncayo
  • Jeanneth Elizabeth Jami Carrera
  • Evelyn Betancourt Rubio

Keywords:

time series models, COVID-19, forecasting, exponential smoothing model, Mean Square Error.

Abstract

Introduction: Time series models are valuable tools to anticipate future COVID-19 values based on historical patterns.

Objective: To predict COVID-19 using time series modelling in patients at a clinic in Ecuador.

Methods: The study was predictive. The study population included patients with a diagnosis of COVID-19 seen at the clinic. The variable of interest was the number of COVID-19 cases, and time-related variables and time series model parameters (alpha, gamma and delta) were used for predictions. A time series model was created and validated using the Ljung-Box test and other goodness-of-fit tests.

Results: The results revealed a stationary R-squared of -2.190 and an R-squared of 0.658, suggesting an adequate ability of the model to capture the trend. The RMSE (Root Mean Square Error) was 8.271, indicating reasonable accuracy of predictions, while the MAPE (Mean Absolute Percentage Error) was 18.970, representing acceptable relative accuracy. The MaxAPE (Maximum Absolute Percentage Error) reached 49.501, and the MAE (Mean Absolute Error) and MaxAE (Maximum Absolute Error) were 4.783 and 14.850, respectively.

Conclusions: The parameters of the exponential smoothing model indicated that past observations had a moderate impact on the level and trend of the model, while seasonality hardly affected the prediction. Significance values suggested that these parameters were not statistically significant in this context.

Downloads

Download data is not yet available.

Published

2023-12-16

How to Cite

1.
Cueva Moncayo MF, Jami Carrera JE, Betancourt Rubio E. COVID-19 prediction using time series modelling in patients in a rural health clinic. Rev Cubana Inv Bioméd [Internet]. 2023 Dec. 16 [cited 2025 Jul. 15];42(2). Available from: https://revibiomedica.sld.cu/index.php/ibi/article/view/3127