1
Department of Physical Education, Farhangian University, P.O. Box 14665-889, Tehran, Iran
2
PhD in Sports Management, University of Tabriz
10.48310/itt.2025.19164.1104
Abstract
Background and Objectives: The application of developing technologies -especially artificial intelligence and deep learning algorithms- in the creation of personalized physical training programs has garnered significant attention, particularly during developmental stages. This study sought to develop and assess an intelligent model for prescribing individualized exercises based on the physical and motor attributes of primary school students. Methods: This descriptive-analytical study was performed on 430 kids aged 12 to 15 in several schools in Tehran. Data were gathered utilizing the Perceived Physical Literacy Instrument (PPLI) and standardized field assessments. Subsequent to preprocessing, the data, encompassing five essential physical components, were analyzed utilizing a multilayer perceptron (MLP) artificial neural network optimized through the ADAM method. The model was developed as a deep learning system to forecast suitable exercises for each individual. Findings: The proposed model effectively recommended personalized exercises, achieving an accuracy of 92.78%, a recall of 85.04%, and a precision of 88.71%. ROC analysis revealed an Area Under the Curve (AUC) of 0.951, signifying a strong ability to differentiate among various workout groups. Alongside the quantitative findings, feedback from students and physical education teachers indicated that 87% of participants were satisfied with the recommended activities, while 83% noted a beneficial effect on their physical health and motivation. Conclusion: The results indicate that deep learning models can function as efficient instruments for creating personalized physical training programs in school physical education. This method improves accuracy, efficiency, and student engagement, while also facilitating ongoing updates and informed decision-making by educators.
kotobi, F. and dehghan banadaki, S. A. (2026). Development of an Intelligent System Based on the ADAM Algorithm for Personalized Exercise Programs for Elementary School Students. Theory and Practice in Teachers Education, (), -. doi: 10.48310/itt.2025.19164.1104
MLA
kotobi, F. , and dehghan banadaki, S. A. . "Development of an Intelligent System Based on the ADAM Algorithm for Personalized Exercise Programs for Elementary School Students", Theory and Practice in Teachers Education, , , 2026, -. doi: 10.48310/itt.2025.19164.1104
HARVARD
kotobi, F., dehghan banadaki, S. A. (2026). 'Development of an Intelligent System Based on the ADAM Algorithm for Personalized Exercise Programs for Elementary School Students', Theory and Practice in Teachers Education, (), pp. -. doi: 10.48310/itt.2025.19164.1104
CHICAGO
F. kotobi and S. A. dehghan banadaki, "Development of an Intelligent System Based on the ADAM Algorithm for Personalized Exercise Programs for Elementary School Students," Theory and Practice in Teachers Education, (2026): -, doi: 10.48310/itt.2025.19164.1104
VANCOUVER
kotobi, F., dehghan banadaki, S. A. Development of an Intelligent System Based on the ADAM Algorithm for Personalized Exercise Programs for Elementary School Students. Theory and Practice in Teachers Education, 2026; (): -. doi: 10.48310/itt.2025.19164.1104
)