Open AccessEpidemic Modelling of COVID-19
Author(s) -
Mayanka Gupta
Publication year - 2021
Publication title -
international journal for research in applied science and engineering technology
Language(s) - English
Resource type - Journals
ISSN - 2321-9653
DOI - 10.22214/ijraset.2021.37439
Subject(s) - support vector machine , regression , polynomial regression , regression analysis , covid-19 , bayesian probability , computer science , ridge , machine learning , elastic net regularization , metric (unit) , statistics , artificial intelligence , polynomial , data mining , mathematics , geography , engineering , medicine , operations management , cartography , mathematical analysis , disease , pathology , infectious disease (medical specialty)
COVID-19 has had a disastrous impact on millions of lives all over the world.199,466,211 confirmed cases of COVID19 and 4,244,541 deaths have been reported to WHO till 4th august. Analyzing the available data and predicting the pandemic trend is important since the situation can be controlled only when there is adequate preparation. Research using epidemiological models helps in analyzing different facets of COVID including infection, recovery and death rate. Predicting the daily increase of cases can help reduce the burden on health care workers and government by aiding them in planning the required resources in advance. Thus, in this project data driven epidemic modelling approach is used. COVID cases of 10 forthcoming days using three modelling techniques namely Polynomial Regression, Bayesian Ridge Regression and Support Vector Machine are predicted. The performance metric used to identify the best model are MSE and MAE. Polynomial Regression is found to have best performance followed by Bayesian ridge regression. Support Vector Machine has a poor performance. Keywords: Epidemic Modelling, COVID-19, Machine Learning, Polynomial Regression, Bayesian Ridge Regression, Support Vector Machine