Decision Making Model for Detecting Infected People With COVID-19

Abstract

The detection of people that are infected with COVID-19 is critical issue due to the high variance of appearing the symptoms between them. Therefore, different medical tests are adopted to detect the patients, such as Polymerase Chain Reaction (PCR) and SARS-CoV-2 Antibodies. In order to produce a model for detecting the infected people, the decision-making techniques can be utilized. In this paper, the decision tree technique based Decisive Decision Tree (DDT) model is considered to propose an optimized decision-making approach for detecting the infected people with negative PCR test results using SARS-CoV-2 antibodies and Complete Blood Count (CBC) test. Moreover, the fever and cough symptoms have been adopted as well to improve the design of decision tree, in which the precision of decision is increased as well. The proposed DDT model provide three decision classes of Infected (I), Not Infected (NI), and Suspected (S) based on the considered parameters. The proposed approach is tested over different patients’ samples in off and real-time simulation, and the obtained results show a satisfactory decision class accuracy ratio that varies from 95% to 100%.