Forecasting of Graduate
Several methods have become increasingly popular in recent
years as reliable tools for predicting enrolment. This section
provides an overview of different approaches used for
School Enrolment using enrolment forecasting.
Using [5] the university's overall enrolment figures from
Model S.Y. 2011–2012 until 2018-2019, the study used the
ARIMA(p,d,q), algorithm model, to forecast the number of
general students enrolled in Cebu Technological University-
Barili Campus for S.Y. 2019-2020 to S.Y. 2024- 2025. The
model with the lowest Akaike Information Criterion (AIC)
I. INTRODUCTION value was evaluated using various p, d, and q values, and it
Forecasting is crucial for effective and efficient planning, was used to make predictions. The simulation's findings
regardless of the situations or time frames involved. Many suggested that the ARIMA (0, 2, 1) approach was
different situations call for forecasting and it may be statistically the most suitable one for predicting university
necessary to make forecasts years in advance or only a few enrolment. The findings also show that enrolment would be
minutes [1]. The application of forecasting to an institution rising during the ensuing school years.
or university is necessary for well-structured strategic
planning of the organization. Improving decision-making
and evaluating potential outcomes can be considered the
benefits of using this tool. The focus of educational data
mining [2] is more frequently on prediction than on
producing precise findings for future goals in which there are
some bright future directions presented by [3] different
prediction models [4] in the field of education. In its ongoing
effort to research the historical, current, and continuing data
correlations for specific scenarios from an educational
environment, higher education institution mining benefits
from prediction [5].
Since predictions make a pertinent conclusion that plays
a crucial role in higher education institutions (HEI)
administration [5], finding an appropriate regression model
[6] that is flexible [7] and will produce realistic predictions
[8] is essential to provide a successful data-driven forecast.
Forecasting a specific outcome uses a modified type of time-
series analysis. There are a number of time series methods to
select based on the characteristics of the time series data. A
good substitute that performs satisfactorily in terms of
prediction is the Auto Regressive Integrated Moving
Average (ARIMA) model [7] or, also known as the Box-
Jenkins technique, which has been considered one of the
most used linear models [9] and can predict future demands
[10] for time series forecasting on the enrolment of the
graduate school. Predicting student enrolment can be
considered as a reference for the future use of the workplace
environment especially on its expenditure and infrastructure.
This also allows the university to specify targets based on the
output data forecasted.
The purpose of this study is to implement the capability
of a subset of regression analysis in the form of ARIMA (p,
d, q) model to forecast the enrolment of the Graduate School
of Central Philippines State University-Moises Padilla
Campus for the next succeeding years. Enrolment data from
each semester of the collegiate year 2016-2017 to the
collegiate year 2022-2023 will be collected and will serve as
the dataset of the study. This study also sought to provide
baseline data on the Graduate School and University
planning office.
II. RELATED LITERATURE
For educational institutions, projecting enrolment is a
crucial endeavour because it allows them to prepare the
resources they will need to meet the anticipated demand.
