| dc.description.abstract | This research will focus on making a prediction of student influx in 2022 for some
of the masters from the faculty of science that are offered at Utrecht University. In this thesis, the number of enrolments in September 2022 for the master
programmes Artificial Intelligence (AINM), Experimental Physics (EXPH) and
Business Informatics (MBIM) is predicted using a Seasonal AutoRegressive Integrated Moving Average with eXogenous regressors (SARIMAX) model. The
deadline for applying for these masters is June first and the number of enrolments is known in September, but it would be best to know the number of
starting students as early as possible so e.g. staff can be hired and schedules
can be made in advance. SARIMAX is first used at the start of May to predict
the total number of applications and this number is then used as external variable for SARIMAX, that also takes the number of applications and enrolments
of previous years into account. The predicted number of enrolling students for
AINM, EXPH and MBIM based on all applications is 76, 15 and 49 respectively.
The 95% confidence intervals, so the expected minimum and maximum number
of enrolments [min, max], are [59, 94], [0, 32] and [32, 66] students respectively
for the three master programmes. While these numbers give an indication of
the number of students that can be expected to show up in September, the
predicted range is still broad and so it might be difficult for Utrecht University
to make decisions based on these predictions. Furthermore, when making influx
predictions before the deadline, using a prediction of the final number of applications, the confidence interval gets even bigger so making precise predictions
before June first is even more difficult. | |
