Analysis of Death Toll from COVID-19 in China with Uncertain Time Series and Uncertain Regression Analysis

Abstract

This paper employs uncertain time series and uncertain regression analysis to model the evolution of the cumulative death toll from COVID-19 in China. Then parameter estimation, residual analysis, forecast and confidence interval are investigated. Uncertain hypothesis test is proposed to determine whether the estimated uncertain statistical models are appropriate. By analyzing the characteristics of the residual plot, the reason of using uncertain statistics instead of probabilistic statistics is explained.