Comorbidity Based Death Risk Prediction in COVID-19 Patients Using Support Vector Machine (SVM)

The covid19 pandemic has hit almost all countries. Covid-19 is a disease caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) which results in many deaths, especially for patients who have comorbidities. By using the machine learning SVM method, predict of the classification between comorbid and death.. The comorbid taken were pneumonia, diabetes, COPD, asthma, insomnia, hypertension, cardiovascular, chronic renal, obesity, and USM. This paper also added with smokers in the feature. The data were taken from Kaggle which is data from the Mexican government from 2020-2021. The SVM method uses a linear kernel and radial basis function to get the F1 value to know which these have the results better. This paper also compares the results of F1 values using other methods such as KNN, Logistic regression, Xboost, decision tree, Random Forest and Multilayer Perceptron and the last, to know the importance feature or which comorbid has the highest death rate using SVM. The Result is SVM uses linear and rbf gets almost the same F1 value. It also same with other methods and the pneumonia has the highest death rate.