Predicting the Specific Student Major Depending on the STEAM Academic Performance Using Back-Propagation Learning Algorithm

The classical educational system in some countries, such as the Arabic countries, depends on the final year’s scores to predict academic performance. In contrast, the STEAM educational system considers the students’ scores for all the studying years alongside the students’ skills and interests to predict academic performance. However, the STEAM educational system predicts the academic performance of students in five to seven general majors regardless of the variant factors that may affect the students’ future careers. Hence, in this research, a seven majors and five factors (SAF) model has been proposed to assign a specific major to every student based on their academic background, interests and skills, and the main influencing factors. The SAF model uses a supervised back-propagation artificial neural network and is trained by a scale-conjugate learning algorithm. The SAF model has the capability to predict a specific major among 17 different majors for every student with a high learning performance (1.4147), plausible error value (-0.1211), and rational number of learning epochs (223).