Prediction Model of Health Insurance Membership for Informal Workers

ABSTRACT :
Background: Health insurance for informal workers is very important to be managed properly by Indonesian government, especially by Health BPJS. Whereas informal workers mostly belong to the category of near-poor people. The current health insurance scheme will depend on personal insurance or assistance from families and surrounding communities. The ability of informal workers is still in dire need of health insurance assistance. This model of assistance can be formulated through certain schemes based on each group’s ability limitations. The purpose of this study is to model predictions of the ability to provide safekeeping funds for health insurance, especially for informal workers in Bogor, through a Data Mining approach. The data used are primary data from a survey of 100 respondents of informal workers in Bogor. The algorithm used to this prediction model is the J48 algorithm. Conclussion:The accuracy of the prediction model for the provision of safekeeping funds is 100% with RMSE values 0.2493 and MEA 0.1243. The accuracy of the prediction model for health facilities selection shows a value of 87%.
KEYWORDS:Data Mining, Prediction, Health Insurance Membership, Informal Workers, J48 Algorithm