A PROPOSED MODEL CAN CLASSIFY THE COVID-19 PANDEMIC BASED ON THE LABORATORY TEST RESULTS

Authors : Şeyma YAŞAR, Cemil ÇOLAK

Pages : 60-63

View : 16 | Download : 9

Publication Date : 2020-12-31

Article Type : Research Paper

Abstract :As reported by the World Health Organization insert ignore into journalissuearticles values(WHO); in March 2020, COVID-19 is a worldwide epidemic. Since the rapid spread of the epidemic harms humans, the need for methods that enable early diagnosis and treatment has increased. Machine learning insert ignore into journalissuearticles values(ML); methods can play a vital role in identifying COVID-19 patients. In this study, the classification algorithms of ML methods insert ignore into journalissuearticles values(CART);, Support Vector Machine insert ignore into journalissuearticles values(SVM-Radial);, K Nearest Neighbors insert ignore into journalissuearticles values(K-NN); and Random Forest are used to determine the best model that diagnoses COVID-19 from the person`s complete blood counts insert ignore into journalissuearticles values(positive/negative);. According to the experimental results, the Random Forest algorithm gives the best predictions in the classification of COVID-19 insert ignore into journalissuearticles values(99.76% of accuracy);. Besides, in the classification of Covid-19, it can be recommended to apply meta-learning algorithms as they can give high predictive results.
Keywords : COVID 19, Machine Learning, CART, SVM, K NN, Random Forest