University Dropout Prediction Using Machine Learning Techniques

Organization
University of Tartu
Abstract
The University of Tartu is interested in better understanding and addressing dropout with the support of data. UT collects and has collected over many years a large amount of data on variables related to student's academic performance, entrance scores, program of study, amount of registered courses and credits, etc. A preliminary decision tree classification model on this dataset suggests that machine learning algorithms can help us identify students at risk of dropout. However, we would like a thesis candidate to develop this model further and benchmark it (using precision and recall metrics) against other machine learning algorithms. The goal of this thesis project is to identify the best machine learning model for predicting the risk of students dropping out of the University of Tartu.
Graduation Theses defence year
2022-2023
Supervisor
Leo Siiman
Spoken language (s)
Estonian, English
Requirements for candidates
Level
Bachelor, Masters
Keywords

Application of contact

 
Name
Leo Siiman
Phone
E-mail
leo.siiman@ut.ee