EFFICIENT PARTICLE SWARM OPTIMIZATION AND FUZZY K-NEAREST NEIGHBOR ALGORITHM FOR DETECTION OF COVID-19

ABSTRACT

With the rapid growth in the population, automatic disease detection is extremely important in medical science. An automatic disease detection framework can assist physicians in identifying a specific disease within a short period of time. It can provide accurate and consistent results at a reduced cost and minimise the death rate. Today, COVID-19 has become one of the most severe and acute diseases that affects the health and well-being as well as the economy of the global population. The critical step to controlling this pandemic and preventing the virus from spreading is the early detection of COVID-19 in the human body. Therefore, an automated disease detection system, as the most accurate, cost-effective, and fastest diagnostic option, is highly required to constrain the exposure and spread of the virus. This paper presents a novel approach for the prediction of COVID-19 disease based on a patient’s symptoms, previous health history, and level of complications. The proposed method uses Fuzzy K-Nearest Neighbor (FKNN) for prediction of the novel COVID-19 in the human body. To achieve better performance, the optimal parameters of the FKNN are obtained using the Particle Swarm Optimization algorithm. The dataset used to evaluate the model was collected by interviewing pneumonia, COVID-19, and non-COVID patients in Kenya, Malawi, and Nigeria. The experimental results indicate that the proposed hybrid system recorded a higher performance in terms of accuracy, sensitivity, specificity, and F1-score as compared to other systems. The proposed system achieved the desired performance using the obtained dataset, which can be further enhanced by integrating more available datasets. The system can assist health experts in detecting the level of stress in humans and providing diagnostic measures.

Keywords: Fuzzy K-Nearest Neighbor, Particle Swarm Optimization, COVID-19, Automatic Disease Detection, Machine Learning