Eureka! Institutional Repository

This master thesis studies the data collected from student examination scores in the Department of Information Technology of TEI of Thessaloniki. Scores from a written examination, are one of the tools for teachers to assess the knowledge acquired by the students. The aim of this work is to find the most important courses of the department, the importance of each one separately an how they associate with each other. Such information might help the teachers to a significant extent to identify weaknesses and shortcomings of the students or potential wrong directions in the teaching process. Early intervention with procedures to be decided by the teacher for students with targeted repetition and feedback, may solve many problems, reduce student failures and increase success rates and increase the knowledge acquired by the students. We processed the score data using machine learning methods that reduce the dimensionality of multivariate data. These methods include principal component analysis (PCA), independent component analysis (ICA) and non-negative matrix factorization (NMF). The results of this study showed that the methods of machine learning could be used in order to identify the importance of each course as well as the key courses in the curriculum.