In December 2019, a novel coronavirus was found in a seafood wholesale market in Wuhan, China. World Health Organization (WHO) officially named this coronavirus as COVID-19. Since the first patient was hospitalized on December 12, 2019, China has reported a total of 78,824 confirmed COVID-19 cases and 2,788 deaths as of February 28, 2020. The COVID-19 has been successfully contained in China but is spreading all over the world. COVID-19 epidemic is prone to disrupt and crumble the existing health- care infrastructures in both the developed and developing world. COVID19 also impacts people’s daily life and country’s economic development. In this paper, we adopt mathematical epidemic models such as Susceptible-Infected-Recovery (SIR), Susceptible-Infected-Recovery-Fatality/Deaths (SIR-F) to sim- ulate the epidemic on the data available for the entire world and future projections on the number of infections, deaths in six specific countries (Italy, France, Spain, Germany, USA, and India) across a time-frame of 7 days, 1 month, 3 months, and 3 years in future. We analyzed the epidemic by extending the SIR-F model with controlled parameters and simulating the behavior on our default case study data. We also fit other mathematical models such as exponential and logistic models to C(t), the cumulative number of positive infections trajectory function. In the latter section, we also used statistical machine learning techniques such as Polynomial regression, support vector machine regression, and simple neural network such as multilayer perceptron to better understand and learns the underlying pattern of the real epidemic growth and the virus proliferation pattern. We found out that the predictions by the logistic model was underreported, i.e, the actual trajectory is more complex than the logistic model. However, we found out that different models found to be better in modeling the pandemic outbreak in respective countries. We also performed data analysis to project the infection, recovery, and death statistics from the real data and also calculated the growth factor of pandemic outbreak in the countries and grouping them respectively. To our future projections and analysis, we found out USA, followed by India are gonna be the most affected countries with each resulting into millions of positive infections cases and deaths.
COVID-19, epidemic modeling, data analysis, SIR/SIR-F modeling, machine learning, polynomial regression, support vector machine, logistic modeling, predictions.
pip install requirements.txt