A Statistical Model Capable Of Predicting Start Of COVID-19 Multiple Waves
The most difficult challenge facing modern scientists is predicting COVID-19 multiple waves. Due to the severity of the accompanying respiratory condition, the so-called COVID-19, the severe acute respiratory syndrome of coronavirus 2 (SARS-CoV-2), spread very swiftly, generating enormous worry at the international level. The first SARS-CoV-2 coronavirus outbreak was observed in late December 2019 in Wuhan, Hubei Province, China, with the COVID-19 sickness spreading fast beyond China's borders. COVID-19 has spread to several nations in just a few months, prompting the World HealthOrganization (WHO) to declare a pandemic on March 12th, 2020. This pandemic has been recognised as one of humankind's most significant severe public health issues in recent history; in less than two years, deaths legally registered around the globe have totalled more than 5.65 million lives, a health catastrophe not seen since the Spanish flu pandemic in the early twentieth century.
COVID - 19 Outbreak Prediction using Machine Learning | Machine Learning Training | Edureka
Using Past Data For Prediction
Existing research in the available literature is concerned with estimating the start date of a local epidemic outbreak of COVID-19, which attempts to infer the beginning date of COVID-19 spread in Western Europe and the Americas, that performed a similar workout for the beginning phases of COVID-19 in Portugal. Then another study examined the initial COVID-19 dynamic behaviour of 10 states in the United States, United Kingdom, Italy, and Spain.
Research published during the first week of March 2022 in the journal "Chaos." by a team of researchers from three Brazilian universities, the Federal University of So Paulo, the Rio de Janeiro State University, and the Federal University of Fronteira Sul, used epidemiological surveillance data, algebraic statistical models, nonlinear regression, and a model selection procedure to infer the systematic behaviour of COVID-19 outbreaks, with a strong emphasis on predicting the potential start date for every infection wave.This approach was shown using COVID-19 data from the city of Rio de Janeiro to identify likely start dates for numerous epidemics in 2020 and 2021.
The primary objective is to propose a strategy that contains generality features and may be employed in typical disease outbreaks by researchers who do not have much experience with epidemiology or complex statistical approaches. This method integrates surveillance data with algebraic multiple waves models, nonlinear regression, and information criteria to produce a basic but realistic statistical model of the underlying epidemic that yields an interval estimation for the referenced infectious wave's start date. The suggested technique is demonstrated using COVID-19 data from Rio de Janeiro, which includes complicated dynamics, many contagion waves, and extremely irregular data.
Starting Dates And Major Events Have A Strong Correlation
As an initial estimate for a regression procedure that aims to calibrate an existing design with six waves, the attributes discovered in each of the waves were used. The close temporal proximity between the plausible period for the onset of community transmission (February 29 or earlier) and the Carnival celebrations in the streets is a sliver of evidence in support of the widely held hypothesis among epidemiologists that the Carnival celebrations in the streets were a critical event in the spread of COVID-19 in Rio de Janeiro. Similarly, correlations with other significant events, such as business reopenings, prolonged vacations, holiday dates, and other celebrations, might be studied.
Conclusion
The precise genesis of new community contagion outbreaks is difficult to pinpoint due to the complicated and poorly understood nature of the disease transmission mechanism. This study proposed a statistical technique capable of characterizing the many waves of transmission observed while also offering an interval assessment of the likely start date for each epidemic.
The results suggest that logistic models with one or more waves may be utilized to generate mathematical explanations in epidemiological situations with complicated transmission dynamics that match well to the facts. Furthermore, they believe that COVID-19 may have begun to circulate locally in Rio de Janeiro as early as February 2020.
