The Enterprise School Readiness Prediction System (ESRPS) Uses Machine Learning to Assess Children's Readiness for Entering Elementary School

This study aims to develop and evaluate the Enterprise School Readiness Prediction System (ESRPS) to predict children's readiness for elementary school using machine learning algorithms. This research employs the Research and Development (R&D) method using Borg and Gall’s model and Instruments include questionnaires, programming tools, performance evaluation metrics, and web/database development tools to ensure the system's validity, reliability, and practical applicability.The research analyzes data from 300 students in various Indonesian cities, focusing on attributes like age, gender, and parental education. The system implements four algorithms: Decision Tree, Random Forest, Naive Bayes, and SVM. Data preprocessing, model training, and hyperparameter tuning were conducted, followed by evaluation using metrics like accuracy and precision. A web-based application was developed for user interaction and deployment. The result showed that the Decision Tree and Naive Bayes algorithms achieved the highest accuracy at 55%, followed by SVM at 50%, and Random Forest at 45%. This suggests that simpler models may be more suitable for the dataset's characteristics. The system also demonstrated the feasibility of practical deployment for educational use. The study concludes that ESRPS effectively uses machine learning to assess school readiness, highlighting the value of data preprocessing and model tuning in enhancing accuracy. Despite moderate accuracy levels, the study confirms the system's potential for aiding educators and parents in supporting children's transition to school.