Determination of kovid-19 transcriptomic biomarkers using machine learning methods

The epidemic caused by the SARS-CoV-2 virus, which emerged in China in December 2019, remains a critical threat worldwide. The SARS-CoV-2 virus causes the deadly disease, covid-19, in humans. Symptoms seen in other viral chest diseases can also be seen in people with Kovid-19 disease. Two important issues can be addressed for Covid-19. These are as follows; First, the patient may not have any symptoms but can still infect other people, and secondly, patients with covid may show the same symptoms as other respiratory infections. For this reason, large-scale screening and studies are needed for the diagnosis of Covid-19. Today, scientists work intensively for the diagnosis and treatment of covid-19. Studies with machine learning methods have an important place in the diagnosis of Kovid-19. Many studies have been done on this subject. In this study, we used Genetic Algorithm, Particle Swarm Optimization Method, Best Priority feature selection methods on blood express data of infected and healthy patients that we obtained in Gene Expression Omnibus (GEO) database for the diagnosis of Kovid-19 with machine learning methods. We reduced the number of cases and then predicted the disease with various classification algorithms (Naive Bayes, Bayesian Networks, k-NN, Random Forest, Logistic Regression, Linear SVM, Radial Based Function SVM, Polynomial SVM, Multilayer Perceptron).