Remarks on a data-driven model for predicting the course of COVID-19 epidemic

TL;DR

This paper introduces a data-driven model for COVID-19 epidemic prediction based on a quadratic log-infected function, providing explicit formulas and parameter relations derived from statistical data analysis.

Contribution

The authors derive explicit formulas for epidemic dynamics using a quadratic log-infected model and validate parameter relations through statistical analysis, enhancing understanding of COVID-19 progression.

Findings

Parameters are linked through validated relations.

The log of infected cases is quadratic in time.

Formulas can be used for data validation and model updates.

Abstract

Norden E. Huang, Fangli Qiao and Ka Kit Tung presented a data-driven model for the COVID-19 epidemic in which the relevant functions depend on a set of seven parameters obtained from a statistical analysis of the available data. These parameters are not independent, they are linked through a set of relations the authors call Main Results which are validated by a statistical analysis of the data. The parameters in questions and the relations between them are not always explicitated by the authors. By given them here their (simple) mathematical formulations all the relevant functions describing the dynamic can be explicitely written down. All the explicit formulas follow from the fact that the log of the number of infected, is a quadratic function of time. The formulas presented here are not themselves approximations - but the parameters they involve are of course statistical quantities derived from the data. These formulas could maybe be of some use either to validate the data, the model itself, to update the model or to find approximations to the relevant quantities.