Application of Machine Learning Algorithms to Predict Alumni Employability Based on Tracer Study Data

Abstract

Tracer study is an important instrument for higher education institutions to evaluate graduate performance and readiness for the labor market. However, the utilization of tracer study data is often limited to descriptive analysis, which lacks predictive capability. This study aims to apply machine learning algorithms to predict alumni employability based on tracer study data. The proposed method consists of data preprocessing, classification modeling using Decision Tree, K-Nearest Neighbor, Support Vector Machine, and Random Forest algorithms, and model evaluation using accuracy, precision, recall, F1-score, and confusion matrix. Experimental results indicate that the Random Forest algorithm outperforms other models by achieving the highest accuracy. Feature importance analysis reveals that soft skills, internship experience, and hard skills are the most influential factors affecting alumni employability, while academic indicators such as GPA contribute less significantly. This study demonstrates that machine learning approaches are effective in enhancing the utilization of tracer study data and can be developed as a computer-based decision support system to improve graduate quality.