Predicting the University Placement Status of University Students Using Artificial Intelligence

Abstract

A university, also known as a higher education institution, is an institution where the highest level of education, research and knowledge is produced. Universities, which are divided into various disciplines, generally consist of units that provide higher education, undergraduate and graduate education. Students who want to attend university after high school are placed in universities according to many criteria such as high school average score, aptitude exams, general exams, language exams, etc. Since there are paid/unpaid institutions in the American university system, it is seen that the student's family status (family income, housing status, etc.) is also an important factor in studying at university.  In this study, it is aimed to predict whether the student will be able to go to university or not by using 4 machine learning models (Decision Tree, Random Forest, K-Nearest Neighbours, Logistic Regression) and an artificial neural network (Multi Layer Perceptron - MLP) methods using the "Go to College" dataset, which is a synthetic and open source 1000 student data.  In the training phase, 5-fold cross validation was used to obtain more accurate results. For a two-state classification problem, 92% accuracy was obtained after training the artificial neural network for 2000 iterations. This value appears to be the best result.