Time series model for the characterization and prediction of the graduation rate at the University of Cartagena

Abstract

For universities, characterizing graduation rates is a key indicator reflecting academic quality and the institution's ability to guide students towards successful entry into the workforce. This paper proposes a time series model for characterizing and predicting graduation rates at the University of Cartagena. Methodologically, the CRISP-DM methodology was adapted into four phases: P1. Business and data understanding, P2. Data preparation, P3. Model construction and evaluation, and P4. Deployment. As a result, various ARIMA models were implemented and evaluated to determine the best fitting model. This model serves as a reference for developing tools to support decision-making by university administrators regarding the number of graduating professionals, in line with educational quality standards.

Keywords

• Graduates
• graduation rate
• predictive model
• time series
• ARIMA