Optimization of the Educational Experience in Higher Education Using Predictive Artificial Intelligence Models

• Published in: RGSA : Revista de Gestão Social e Ambiental Vol. 18; no. 5; p. e07111
• Main Authors: Gallastegui, Luis Miguel Garay, Forradellas, Ricardo Reier
• Format: Journal Article
• Language: English
• Published: 24.05.2024
• ISSN: 1981-982X; 1981-982X
• DOI: 10.24857/rgsa.v18n5-104

Aim: This study investigates the application of machine learning-based predictive models in university education, in order to improve student experience and satisfaction, evaluating the effectiveness of these tools in a modern educational context.   Theoretical Framework: The research analyzes the strategic transformation in higher education, driven by digitalization and the evolving expectations of students and the labor market. The crucial role of AI-based predictive models in this change is explored.   Methodology: A methodology combining opportunity identification and business case development in educational settings is employed. The approach focuses on business design and experimental machine learning techniques, emphasizing model accuracy, evaluation of the costs of inaccurate predictions, and ethics in data manipulation.   Results: The predictive models achieved 95.7% accuracy in predicting student satisfaction, showing a significant positive correlation between teaching personalization and student satisfaction. These results highlight the ability of the models to influence educational decisions that improve the student experience and underscore their adaptability to specific learning needs, thus contributing to a more personalized and effective education.   Discussion: The importance of balancing the adoption of advanced technologies with the maintenance of a student-centered pedagogical approach is emphasized. The methodology used to identify modeling heuristics highlights how strategic decisions can guide the technical development of AI models, ensuring that solutions are not only innovative, but also aligned with educational needs.   Research Implications: The approach taken suggests that the application of predictive models has the potential to radically transform teaching and learning, aligning them with the demands of the digital future.   Originality/Value: The proposed machine learning model is revealed as an effective tool to identify areas for improvement in university education. Its high accuracy in classifying students provides a unique perspective on how to improve educational quality and student satisfaction, emphasizing the importance of addressing individual needs for educational improvement.