NEW STUDENT PREDICTION SYSTEM USING THE LONG SHORT TERM MEMORY METHOD BASED ON TIME SERIES DATA

Abstract

Education plays a crucial role in improving the quality of human resources. One of the indicators of a primary school’s success is the stability of new student enrollment each year. However, fluctuations in student numbers often become obstacles in school planning, including facilities, teaching staff, and budgeting. This study aims to develop a prediction system for new student enrollment at SDN 1 Kedungjaya using the Long Short Term Memory (LSTM) method with a time series data approach. The data used consist of new student enrollment records from several previous years, followed by preprocessing, model training, and evaluation using Mean Squared Error (MSE). The results show that the LSTM model is capable of predicting new student enrollment with relatively low error rates. The system is implemented as a web-based application, enabling schools to utilize it as a decision support tool for future enrollment planning.