Forecasting Elderly Well-Being through Decision Tree Modeling Techniques: Integrating Google Maps for Community Engagement in Bang Jakreng, Samut Songkhram Province, Thailand

Abstract

This research is a quasi-experimental study focusing on the application of Decision Tree modeling techniques. The objectives are as follows: (1) to study the well-being data of the elderly in the Bang Chakreng community, Samut Songkhram Province; (2) to develop a predictive model and apply it to generate community health statistics using decision tree modeling; and (3) to incorporate Google Maps for managing data on community engagement. The sample group consists of 174 elderly individuals from the Bang Chakreng community, with the analysis focusing on four factors: overall health (Hel.), hygiene (Hyg.), environment (Env.), and economy (Eco.). The data was divided into training and testing sets using K-folds cross-validation and percentage split techniques. The results indicate that the optimal model yielded an accuracy rate of 91.40%. The predictive data can be utilized to plan and improve the well-being of the community moving forward.