Machine Learning Based Selection of Incoming Engineering Freshmen in Higher Education Institution

Abstract

The Accrediting Agency of Chartered Colleges and Universities of the Philippines recommends through the university testing unit, a system to interpret and analyse entrance test results that may help direct and guide students in choosing a Baccalaureate degree to take in the college. While the present system of manually evaluating each of the freshman applicants is used, there is a need to adopt technological tools for faster and accurate analysis. Thus, the study presents machine learning methods of classifying freshmen applicants if they are qualified or not in the college of engineering and architecture. Specifically, determining if a freshmen applicant may succeed in the five engineering program at university. The study used classifiers such as Decision Tree, K-Nearest Neighbor (KNN), Decision Tree, and Support Vector Machine (SVM). A cross-validation of ten-fold model was used better accuracy of classifiers. The predicted models performed well, however, the Decision Tree classifier outputs a higher average accuracy and F1measure. The result shows that the classifier accurately classifies qualified and non-qualified engineering freshmen for program acceptance.

Predictive approach to the degree of business process change

“…We used cross-validation in ML to estimate the skill of machine learning models on a new dataset [22]. We can see that the result displayed in the Table IX…”