<![CDATA[Students’ academic achievement is one of the important indicators used to assess the success of the educational process. In addition to being influenced by cognitive factors, academic achievement is also affected by psychological factors, one of which is self-esteem. This study aims to analyze the application of deep learning models in predicting students’ academic achievement based on self-esteem in several elementary schools in Medan City. The study employs a quantitative approach with a predictive design. The research data were obtained through self-esteem questionnaires, documentation of students’ academic scores, and observations of learning conditions in schools. The collected data were then organized into a dataset and divided into 80% training data and 20% testing data. Furthermore, the data were analyzed through the stages of preprocessing, normalization, feature selection, and modeling using deep learning. The results of the study indicate that self-esteem has a fairly strong contribution in predicting students’ academic achievement. The proposed deep learning model also demonstrated good performance, with an accuracy of 91.80%, precision of 92.70%, recall of 93.50%, and F1-score of 92.40%. These findings show that the integration of psychological factors and artificial intelligence technology can be an effective approach in supporting educational evaluation and assisting schools in identifying students who require earlier intervention.]]>
Copyrights © 2026
