Using XG Boost and Random Forest Classifier Algorithms to Predict Student Behavior

Abstract

Students study in an online environment, the problems relate to reaction based on evaluation of student’s performance and students’ skills to understand the student behavior. In this paper, for students in an online environment, techniques for connecting the students’ skills and the online reactions about behavior via their evaluation are considered. An example about students from a Brazilian University of an introductory class of Algorithms for explorative data analysis is applied, an instrument for XGBoost analysis and RandomForestClassifier. A base for evaluation of student achievement is the analysis of behavior. This idea is based on studies that discussed the use of social features in the actual classroom of the project.