교육 분야에서의 랜덤포레스트 알고리즘 활용 연구 동향 분석

Abstract

With the growing importance of artificial intelligence and related technologies, there has been a surge in educational research cases utilizing AI. Among them, the use of the random forest algorithm, which originated in educational evaluation research, has seen a consistent rise. In this study, a trend analysis was conducted on 72 studies using the random forest algorithm published in KCI-listed journals from 2015 to May 2023. The research analysis framework was divided into five main areas: 'utilization purpose, research topic, research subject, analysis method, and utilization of results'. While the random forest algorithm has been employed in various fields to draw educational insights, there were instances that required improvement such as the use of limited or imbalanced data, misuse of model performance metrics, and non-disclosure of performance indicators. Based on these findings, recommendations were made for the use of the random forest in the educational sector to enhance reliability and validity. The significance of this study is underscored by the fact that its suggestions can be applied not only to the random forest but also to other machine learning algorithms.