Open AccessAn Analysis of the COVID-19 Epidemic in Japan Using a Logistic Model
Author(s) -
Kuniaki Miyamoto
Publication year - 2021
Publication title -
journal of disaster research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.332
H-Index - 18
eISSN - 1883-8030
pISSN - 1881-2473
DOI - 10.20965/jdr.2021.p0012
Subject(s) - epidemic model , covid-19 , pandemic , population , logistic function , logistic regression , field (mathematics) , limit (mathematics) , simple (philosophy) , econometrics , statistics , demography , computer science , operations research , mathematics , medicine , sociology , philosophy , epistemology , mathematical analysis , disease , pathology , infectious disease (medical specialty) , pure mathematics
The COVID-19 pandemic has been persistent. For example, the number of infections increased exponentially since mid-September in Europe. The SIR, among other models, is used to examine and detail such epidemics and the changes they bring about. However, the application of the models requires that we fix the parameters to govern the processes; it is difficult to set them up appropriately, especially during new epidemic cases such as COVID-19. If we can limit the purpose of analysis to understand current epidemic situations, then it would be better to use simple models and limit the number of parameters. The logistic model is one of such suitable models, which can reflect the basic characteristics of an epidemic to provide information on the state and tendency of the epidemic based on little information. This research uses daily cases, deaths, and recoveries to analyze the epidemic and derives interesting results. The first wave of the epidemic, which ran from March to May, almost complies with the logistic model. In the case of the second wave, since mid-June, the results show that the rising phase has characteristics similar to those of the first wave. However, the phase of decline has different characteristics. Currently, in mid-October, it is almost in a state of equilibrium. This result means that the data used in this analysis show some characteristics of the statistical population of the “epidemic field.” However, while we consider the fact that infected persons must be isolated and hence removed from the “field,” it is suggested that the number of infected and recovered persons must be significantly larger than that of the reported cases. Nevertheless, it is difficult to evaluate the statistical characteristics of the “epidemic field” using the data, as they are not the results of “random sampling.”