This study introduces a novel approach using machine learning to help students choose their specialization. Our goal is to address the impact of factors such as parental finances and coordination office decisions on college admissions, which often override student preferences. By analyzing a comprehensive dataset from 2011 to 2023, we examine the behavior of all enrolled students across different programs. Using their attributes, we develop a new predictive model to suggest specializations or programs that align with a new student's profile, drawing insights from similar profiles in the historical dataset. In this study, we implement four classifiers to identify the most informative attributes that play a major role in determining and evaluating student selection or scholarship grants based on their program performance. The classifiers used are classifier_subset_evaluator, information_gain_ranking filter, attribute evaluator, and wrapper_subset_evaluator. Each classifier produces different outputs. We found that HIGH_SCHOOL_YEAR ranked at 0.617542, HIGH_SCHOOL_NAME at 0.569052, and HIGH_SCHOOL_SCORE at 0.417498. These top three attributes could play a significant role in determining if the selected program will be a suitable fit for the student based on their profile. The evaluation metrics, such as TP Rate, FP Rate, Precision, Recall, F-Measure, ROC Area, and PRC Area, have achieved high scores, with all metrics scoring 1.000 for the “graduated” class. Given that an AI application generated these outcomes and recommendations, we believe they are valuable for universities, ministries, families, and students.