Comparative analysis of machine learning models for concentration level prediction in e-learning

Abstract

Concentration level plays an important role while accomplishing cognitive tasks. Although there are several methods for predicting concentration levels, the majority of the methods now in use for assessing concentration levels rely on labor-intensive and time-consuming human coding. In this work, learners’ performance and involvement in online assessments are utilized to categorize the concentration level using a publicly accessible dataset. The goal of the project is to identify the best machine learning (ML) model for concentration level prediction. Six distinct ML classifiers are employed for comparative analysis in this context. When compared to alternative classifiers for the assessment of concentration level, the eXtreme gradient boosting (XGBoost) model has the best precision (80.87%). These findings are highly promising and have implications for the future development of automated concentration level analytic tools for online learning.