Determining the effective factors in engineering education and predicting the increase of academic years with multi-criteria decision making and data mining approach (Artificial Neural Network)

Considering the role of universities in engineering education, it is necessary to study the status of the educational system and its strengths and weaknesses in order to improve the process of engineering education. In this research, the factors affecting the increase of academic years of engineering students and the status of students who do not complete their studies in the authorized years have been investigated. At first, the effective indicators on increasing the academic years of engineering students were identified and prioritized using the AHP technique. The ranking results show the average of basic courses, the average of main courses, the average of general courses, the number of conditional semesters, the average of optional courses and the number of units dropped have the most impact on increasing the academic years of engineering students. Then, a model was provided to predict the increase in academic years according to the educational status of engineering students using an artificial neural network. According to the results of the neural network, the factors of the number of units dropped, the average of the main courses, the average of the basic courses, the number of conditional terms, the duration of marriage and the average of the high school math and physics courses have the greatest impact on increasing academic years. Finally, by comparing the results of AHP method and neural network, the average factors of basic and main courses, the number of conditional terms and the number of units dropped in both methods were identified as the most effective factors that should be paid more attention while studying engineering students.