Indonesia

Abstract

This study aims to assist the Gereja Batak Karo Protestan (GBKP) Km.7 Padang Bulan Medan in managing its finances more efficiently, particularly in addressing fluctuations in the congregation size that affect weekly income. The method used is multiple linear regression to predict church income. The independent variables in this study include the number of active congregants and weekly tithing, while the dependent variable is the total church income. The data analyzed is historical data from 2020 to 2023. The analysis results using Python show that this multiple linear regression model has a Mean Absolute Percentage Error (MAPE) of 16% and 12.90%. These MAPE values indicate a fairly good prediction accuracy, with low prediction errors. A lower MAPE signifies that this model is reliable enough to be used for church financial planning. Thus, this method can assist the church in preparing more accurate and efficient financial planning based on historical data.