Time Series Analysis on Enrolment Data: A case in a State University in Zamboanga del Norte, Philippines

Abstract

Changes in enrolment would result to many problems such as shortage in human resource and infrastructure. Using Prophet to forecast future student numbers will aid administrators in effectively allocating resources and making future decisions. The data used in this study is the entire population of college students enrolled in Jose Rizal Memorial State University - Tampilisan Campus from 2000–2022. Data shows a fluctuation in enrolment data but significant increase is observable in A.Y. 2013–2014 up to A.Y. 2015–2016 and A.Y. 2018–2019 up to 2021–2022, respectively. Likewise, data shows a seasonal decrease of number of enrolees in the second semester in comparison to first semester in every academic year. Further, results during the training phase in terms of root mean square error (RMSE) and coefficient of determination (R2) of the different forecasting models trained using different enrolment data and Prophet shows that model trained using BS Business Administration (BSBA), BS Agriculture (BSA), and BS Criminology (BSCrim) dataset attains the top three (3) smallest RMSE result of 15.51 and 17, and the top three (3) highest R2 value of 0.97 and 0.95, respectively. On the other hand, model trained using consolidated enrolment data attains an RMSE of 36.7 and a R2 score of 0.87. Based on the findings, different models attain varied results; however, there are models which attain higher degree of accuracy as depicted in the RMSE and R2. This indicates that predicting enrolment data using those models with higher accuracy is similar to real data thus it is viable in predicting future values. The researcher assumes that this study may be implemented and incorporated into current school and university information systems. Further, other mathematical models may be incorporated into the current model to improve forecast accuracy.