Expert System to Model and Forecast Time Series of Epidemiological Counts with Applications to COVID-19

Abstract

We developed two models for real-time monitoring and forecasting of the evolution of the COVID-19 pandemic: a non-linear regression model and an error correction model. Our strategy allows us to detect pandemic peaks and make short- and long-term forecasts of the number of infected, deaths and people requiring hospitalization and intensive care. The non-linear regression model is implemented in an expert system that automatically allows the user to fit and forecast through a graphical interface. This system is equipped with a control procedure to detect trend changes and define the end of one wave and the beginning of another. Moreover, it depends on only four parameters per series that are easy to interpret and monitor along time for each variable. This feature enables us to study the effect of interventions over time in order to advise how to proceed in future outbreaks. The error correction model developed works with cointegration between series and has a great forecast capacity. Our system is prepared to work in parallel in all the Autonomous Communities of Spain. Moreover, our models are compared with a SIR model extension (SCIR) and several models of artificial intelligence.

Substation Equipment Spare Parts’ Inventory Prediction Model Based on Remaining Useful Life

“…Next, we make certain variables labeled and also define a new variable named life course. e textual variables z i (1) to z i (5) are labeled with the regular numerical values and are denoted as x i (1) to x i (5).…”

“…e expert meeting method and the Delphi method are less practical and reliable due to their reliance on expert experience. However, this qualitative model, combined with other methods, has an excellent performance in forecasting [5][6][7]. With the development of computers and the demand for short-term and medium-term forecasting techniques, quantitative forecasting models have taken the dominant position, such as gray forecasting models [8,9], linear regression [10] and nonlinear regression [11], and autoregressive moving average (ARMA) models [12][13][14].…”

Nonlinear Combinational Dynamic Transmission Rate Model and Its Application in Global COVID-19 Epidemic Prediction and Analysis

“…Tang et al [6] proposed that the risk of secondary outbreaks can be effectively reduced by intermittent population mobility and effective isolation of infected people in the floating population based on a novel stochastic discrete transmission model. In the meantime, the data-driven statistical models were also widely used in the prediction and analysis of COVID-19 epidemics, including function fitting models [7][8][9], machine learning [10,11], deep learning [12,13], and time series models [14,15].…”