Chronic Disease Classification for Healthcare Facility Recommendation System

Chronic diseases are a major cause of the global burden of disease, causing widespread death and disability worldwide. However, the underdevelopment of medical facilities in Indonesia has become a serious public health problem. A recommendation on the sequence of construction of health facilities using a chronic disease classification model is given and can be used by the government to allocate resources to build healthcare facilities in areas of greatest need to solve this problem. This study focuses on developing a chronic disease classification model using machine learning with a dataset from the Social Health Insurance Administration Body (BPJS). This dataset includes advanced referral facilities (FKRTL) data, non-capitation first-level facilities (FKTP) data, and membership data. The model will be used to make recommendations in order of health facility development. Research steps include data preprocessing, feature extraction, generation and evaluation of chronic disease classification models, and making recommendations on the development sequence of healthcare facilities. The testing scenarios were performed by comparing different models using test data, testing the confusion matrix as well as recall and f1-score to determine the best model. The results show that the decision tree model with feature selection has a recall of 98.127% and an f1-score of 86.776%, which is the best model to recommend on the order of development of medical facilities.