Bed and Revenue Forecasting in Nursing Home Management System

Abstract

Data management in nursing home typically stored in manual filing systems and hardcopy records. Despite the the adoption of digital systems has begun, the information gathered often remains underutilized, and fails to contribute to nursing home operations. Therefore, this study proposes the integration of predictive analytics to analyse and predict the need for bed capacity and revenue projections to determine the sustainability of nursing homes. This study utilized historical data of bed occupancy and revenue to train two forecasting models, ARIMA and Prophet, which are compared and evaluated based on metrics such as r2_score, mean squared error, and mean absolute error. The results show that the Prophet algorithm outperforms ARIMA in both bed and revenue forecasting. The system is implemented with a user-friendly web interface that allows users to input the date range for forecast and retrieves the forecast result from the backend. The proposed system provides nursing home managers with valuable insights into the future trends of bed occupancy and revenue, enabling them to make informed decisions and better manage their resources.