<![CDATA[Patient visit numbers at the Aek Kota Batu Community Health Center (Puskesmas) fluctuate over time; therefore, a method is required to predict visit volumes for future periods to inform healthcare service planning. This study aims to analyze historical patient visit patterns and develop a prediction model using linear regression. Monthly patient visit data from 2025 to 2026 were processed using the Knowledge Discovery in Databases (KDD) framework, comprising data selection, preprocessing, transformation, data mining, and evaluation. Analysis was conducted using POM-QM for Windows software, incorporating five independent variables based on time-series lags (lag-5 through lag-1). The results demonstrate that linear regression can generate a prediction model based on the relationship between historical data and future patient visit volumes. Model performance was evaluated using R², MAE, RMAE, MSE, RMSE, and MAPE metrics to assess prediction accuracy. The resulting model can serve as a decision-support tool for planning healthcare personnel, facilities, and services at the Aek Kota Batu Community Health Center, thereby enhancing the effectiveness and efficiency of service delivery.]]>
Copyrights © 2026
