ENBIS: European Network for Business and Industrial Statistics

Optimal Admissions in Higher Education Using a Combined Goal Programming and Artificial Neural Network

23 September 2009, 10:25 – 10:45

Abstract

Submitted by

S.Mojtaba Sajjadi

Authors

S.M.Sajjadi,Yashar Entezar Ghofran, S.M.Asadzadeh

Affiliation

Azad University, Branch of Najafabad

Abstract

In this paper, we propose a combined Goal Programming (GP) and Artificial Neural Network (ANN) approach to mathematically determine optimal number of students that can be admitted to six different educational groups (engineering, medicine, human science, art, agriculture, and basic sciences) in six different university groups (Governmental non-medical schools, Governmental medical schools, Azad University, applied scientific Institutes, Payame Noor university, and non-governmental non-profitmaking universities) for three degrees (BSc, MSc, and PhD) in Iran in time period 2009 - 2012. The main contribution of the paper is two folds. i) We employ ANN to estimate the economic demands of higher education (HE) in different educational groups. Also, the social demand for HE that almost other studies in this respect fail to embed it into the modeling, is estimated by modeling the demand process by ANN. ii) Using information generated by the aforementioned two neural networks models, we develop a dynamic linear goal programming model to determine the optimal values for variables of our model i.e. Xijkt that denotes the number of students that can be admitted for degree i in educational group j in university group k in the time period t. Data used is about past 20 years of Iranian educational system (1988-2007).

View paper

Return to programme