Data modeling COVID-19 patients in Thailand: data mining techniques

Abstract

This study aimed to investigate the characteristics of COVID-19 patients in Thailand and develop a data model for analyzing these characteristics. A total of 1,888,941 cases from the Thailand Department of Disease Control website from January 12, 2020, to October 29, 2021, were analyzed, and 20,110 cases were selected for further analysis. The two-step cluster analysis method was used to cluster the data according to four variables: nationality, occupation, patient type, and risk groups. The results showed the presence of three groups of COVID-19 patients. Group 1 consisted of 5,883 workers in trade and service occupations who had contact with the public and were either Thai nationals or from abroad. Group 2 was the largest cluster, consisting of 7,420 migrant workers classified as foreigners and working in industrial settings. Group 3 consisted of 6,870 cases of indirect transmission, with individuals in this group infected through close contact with family members or individuals in the first two groups. This clustering approach offers valuable insights for pandemic management, aiding in identifying high-risk groups and developing tailored interventions. In future outbreaks with similar characteristics, such as airborne transmission, contact infection, or super spreader events, our model can serve as a valuable tool for devising effective management plans and countermeasures. In conclusion, this study emphasizes the significance of cluster analysis in understanding the dynamics of COVID-19 transmission and highlights its potential for informing effective pandemic management strategies. It also outlines promising avenues for further research to enhance our knowledge of COVID-19's impact on specific populations and inform future public health efforts.