AN EXPLORATORY DATA ANALYSIS WITH MACHINE LEARNING PREDICTION OF NOVEL CORONAVIRUS

5 ABSTRACT This project described different aspects of corona virus disease (COVID-19), presented a visualization of the spread of the infection, and discussed the potential applications of data analytics on this viral infection. Firstly, a literature survey was done on COVID-19 highlighting a number of factors, including its origin, similarity with previous corona viruses, and transmission capacity. Secondly, data analytics is applied to a dataset of Johns Hopkins University to find out the spread of the viral infection. The worldwide increase in the number of confirmed cases over time is modeled here using a linear regression model, polynomial regression algorithm with degree 5, Support Vector Machine model, Holt’s linear Model, and Holt’s Winter model. Fourthly, classification algorithms of the four models are applied on a dataset of 76,997 samples provided by Kaggle. The models were used to predict the confirmed cases in 7 days. It is observed that Holt’s winter model gave the most accurate prediction amongst other models as it tends to have been the closest to value in accordance to the root me