C4.5 Decision Tree Algorithm and Linear Regression in Guidance and Counseling Decision Support System

Abstract

This study employs a combination of quantitative and experimental approaches to explore the relationship between family dynamics, psychological test results, and academic performance among students in specific courses. Quantitative data collection was conducted using existing student profiles from the guidance office, while an experimental research method utilized the C4.5 algorithm to determine these relationships. The participants were students enrolled in various courses over a specific academic year. The decision tree generated by the C4.5 algorithm was used to predict academic performance based on family dynamics and psychological test results. The study also employed Simple Linear Regression to identify the predictors of academic performance. The accuracy of the predictive model was tested using a confusion matrix, which yielded an accuracy rate of 97%. Additionally, a correlation matrix and regression analysis were performed to identify significant correlates and predictors of academic performance. The findings of this study suggest the potential for positive changes in guidance services, particularly in identifying predictors of academic performance and informing decision-making processes. These findings contribute to our understanding of schooling outcomes in relation to family dynamics and parental migration status.