Predictive Analysis On Recovery of Covid-19 Patients in Singapore by Using Data Mining Techniques
Keywords:Decision Tree, K-Nearest Neighbors, Naïve Bayes, COVID-19
Whilst the researchers are actively searching for infections and recovery data across countries actively, the information of patients' recovery on COVID-19 disease is poorly recognized. There is a lot of uncertainty of the mild or asymptomatic COVID-19 cases in the clinical presentation that may never present to healthcare services. The purpose of this study is to develop a predictive model for estimating a COVID-19 infected patients’ recovery by using the dataset of COVID-19 patients in Singapore from January to February in 2020. Data Mining techniques such as Decision Tree, K-Nearest Neighbors and Naïve Bayes were applied in this study. The models help to predict which causes of infection cases are more likely to recover from COVID-19; the maximum and minimum number of days for the patients to recover, and the recovery rate of the different age groups. The result of this research shows the Decision Tree model has been recorded as the most efficient data mining technique in predicting the recovery of COVID-19 patients in Singapore with the greatest percentage of accuracy of 78.95% among other predictive models. In future research, it is suggested to use Malaysian data with more relevant attributes and input samples so that it will reflect a better performance to the government about the trend of the COVID-19 patient recovery.