This study demonstrates effective k-means clustering and naïve bayes classification methods on tracer study data, indicating improved graduate quality assessment.
Key Points
The study achieved an accuracy of 95.24% using the k-means clustering and naïve bayes techniques.
K-means clustering grouped graduates into three clusters based on the most optimal K value identified by the silhouette score.
The use of the SMOTE-ENN method helped balance the dataset, enhancing model effectiveness.
The approach can serve as a decision-making tool in improving the quality of higher education.