Patient'S Hospital Choosing Behavior - Comparison of Results between Econometric Modeling and Models of Machine Learning Mechanism

While the National Health Insurance in Taiwan protected the freedom of choice for each citizen during accessing healthcare services, people in Taiwan commonly make their own decisions on which institute or which doctor to visit. However, the public did not possess enough knowledge to choose suitable facilities for themselves, instead of consider individualized needs or condition, a lot of people make such decision based on unreasonable information, such as recommendation from others, reputation of the institute, hospital levels of the facility, or merely choose the institute near by. This had been pointed out as ineffective ways of utilizing healthcare resources. With proper tools or information, the public can make better decision, which will results in better experiences and better quality of care. The aim of this research is the to use machine learning mechanism to analyze the choice of patients and build a recommendation system for patients to use. This study initially collected the affecting factors of the patients, the care providers, and the incidents. The factors will be further analyzed with feature selection and classification technique to find the data model that explicitly interprets patient behaviors, and generate recommendations for them as a result. The constructed data model will be compared with existed economic models, and through the comparison and refining the model, a better and more accurate recommendation system is expected.

Project ID:PF10907-1946
External Project ID:MOST109-2410-H182-006