<![CDATA[Background: The scholarship selection process often faces problems because of the large number of applicants that mustbe evaluated based on various criteria. This makes the manual selection process take a long time and can lead to subjectivedecisions.Objective: This research aims to build a prediction model for scholarship eligibility using the Logistic Regression algorithmwith the Bagging technique.Methods: This study uses the CRISP-DM methodology with a dataset of 2,042 student records. The attributes includeGPA, parents’ income, number of dependents, and organizational participation. The data is divided into 80% for trainingand 20% for testing to build and evaluate the model.Result: The research results show that applying the Bagging technique increases accuracy from 82.89% to 83.37%, with aROC AUC of 0.9223. The model was then implemented into a web-based prediction system using Flask.Conclusion: These results show that Logistic Regression with the Bagging technique can be used to support a scholarshipselection process that is faster, more objective, and more consistent. ]]>
Copyrights © 2026
