posted on 2016-07-01, 00:00authored byMaryam Teimoori
Exploring the study paths of students in higher education is crucially important in order to maximize their likelihood of academic success. The study path of a student is the sequence of courses that the student takes in order to graduate. The main focus of this dissertation is to identify and analyze the study paths of Mechanical Engineering (ME) undergraduate students at the University of Illinois at Chicago (UIC). Using students’ academic records, several machine learning techniques were applied in order to explore the behavior of students from admission to graduation. The major contributions of this thesis are as follows. First, it determines whether the current ME university approved study path is consistently followed by students. Using clustering methods, the study paths that are actually followed by students are derived. The graduation time and the Grade Point Average (GPA) of students in each study path are measured. Furthermore, classification techniques are used to predict the study path of new incoming students. Second, a new index in defined in order to calculate the overlap ratio of the courses taken by ME students. This index is used to measure students’ tendency to take any given pair of courses in the same semester.
The contributions of this thesis can support both university students (through advising) and their academic departments (through scheduling). Students can benefit from receiving an enhanced advising (based on selected study path) that may result in a shorter graduation time and a higher GPA in each semester. Departments can achieve a more accurate course enrollment prediction and hence, provide a better course scheduling. Moreover, based on the study path clusters, students’ graduation time can be projected which can help departments in assigning their classroom and teaching resources more efficiently.