BACKGROUND
Statistical predictions are useful to predict events based on statistical models. The data is useful to determine outcomes based on inputs and calculations. The Crow-AMSAA method will be explored to predict new cases of Coronavirus 19 (COVID19). This method is currently used within engineering reliability design to predict failures and evaluate the reliability growth. The author intents to use this model to predict the COVID19 cases by using daily reported data from Michigan, New York City, U.S.A and other countries.
The piece wise Crow-AMSAA (CA) model fits the data very well for the infected cases and deaths at different phases during the start of the COVID19 outbreak. The slope β of the Crow-AMSAA line indicates the speed of the transmission or death rate. The traditional epidemiological model is based on the exponential distribution, but the Crow-AMSAA is the Non Homogeneous Poisson Process (NHPP) which can be used to modeling the complex problem like COVID19, especially when the various mitigation strategies such as social distance, isolation and locking down were implemented by the government at different places.
OBJECTIVE
This paper is to use piece wise Crow-AMSAA method to fit the COVID19 confirmed cases in Michigan, New York City, U.S.A and other countries.
METHODS
piece wise Crow-AMSAA method to fit the COVID19 confirmed cases
RESULTS
From the Crow-AMSAA analysis above, at the beginning of the COVID 19, the infectious cases did not follow the Crow-AMSAA prediction line, but during the outbreak start, the confirmed cases does follow the CA line, the slope β value indicates the pace of the transmission rate or death rate in each case. The piece wise Crow-AMSAA describes the different phases of spreading. This indicates the speed of the transmission rate could change according to the government interference, social distance order or other factors. Comparing the piece wise CA β slopes (β: 1.683-- 0.834--0.092) in China and in U.S.A (β:5.138--10.48--5.259), the speed of infectious rate in U.S.A is much higher than the infectious rate in China. From the piece wise CA plots and summary table 1 of the CA slope βs, the COVID19 spreading has the different behavior at different places and countries where the government implemented the different policy to slow down the spreading.
CONCLUSIONS
From the analysis of data and conclusions from confirmed cases and deaths of COVID 19 in Michigan, New York city, U.S.A, China and other countries, the piece wise Crow-AMSAA method can be used to modeling the spreading of COVID19.