Predictive Models of Hospital Readmission Rate Using the Improved AdaBoost in COVID-19
Arash Raftarai, Rahemeh Ramazani Mahounaki, Majid Harouni, Mohsen Karimi, Shakiba Khadem Olghoran
Abstract
In 2019, an unknown virus called COVID-19 was identified, which affected all aspects of life in the world. With the outbreak of this virus, the need for health services became more and more important. Because of the limited resources to provide health services due to the epidemic of the disease, the issue of improving the quality of services provided to patients with COVID-19 was considered by governments, organizations, and health institutions. Improving service quality in the health industry is one of the paramount parameters. Along with increasing the service quality, increasing the cost should also be controlled. Readmission rate reduction is one of the parameters of health quality improvement. Readmission rate prediction is a complex process. Machine learning and data mining-based methods and algorithms have appeared successful for readmission rate prediction. A new predictive model of readmission rate based on the improved AdaBoost ensemble classifier is proposed in this chapter. The proposed model is based on machine learning techniques and intelligently combines three classifiers in an ensemble classifier. Obtained results have been evaluated by accuracy 91.61%, sensitivity 95.80% and positive predictive values (PPV) 90.25% and negative predictive values (NPV) 89.31% and also compared with basic classifiers. The obtained results showed the superiority of the proposed model.