Complex competencies for leader education: artificial intelligence analysis in student achievement profiling

Abstract

Future education requires fostering high-level competencies to enhance student talent, and artificial intelligence (AI) can help in profile analysis. The aim was to determine the variables that predict the GPA of students in the ‘Leaders of Tomorrow’ program through an integrated methodology of data analytics, machine learning modeling, and feature engineering in order to generate knowledge about the application of AI in social impact programs. This research focused on 466 graduates of a ‘Leaders of Tomorrow’. A regression analysis was performed to model the relationship between the dependent variable and multiple independent variables. The findings revealed: (a) Analysis of variance (ANOVA) demonstrated exceptional model fit for predicting ‘student.term_Grade Academic Performance (GPA)_program’ with an R-squared of 0.999; (b) Visual analysis showed that significant variables like age and origin-school Grade-Point Average (GPA) affect term GPA; (c) Kendall tau correlation revealed a positive correlation of origin-school GPA with term GPA and a slightly negative one with age; (d) Support Vector Machine (SVM) regression aligned actual and predicted GPAs closely, indicating high accuracy; and (e) Recursive Feature Elimination (RFE) identified ‘student_originSchool.gpa’ as the most predictive feature. This study is intended to be of value to academic communities interested in enhancing the academic profiles of students with complex competencies, as well as communities interested in applying AI in education for predictions that contribute to trajectories for training.