Prediction of Learning Performance in Online Course Based on Linear Regression Model: IEEE ITAIC(ISSN:2693-2865)

Abstract

This paper takes the online learning behavior and performance data of online courses as the research object, analyzes the factors that affect the course performance in online learning behavior, and establishes a performance prediction model. First, the influencing factors of course achievement are analyzed by calculating the correlation coefficient between learning behavior features and course achievement; then, a linear regression performance prediction model is constructed by using single or multiple learning behavior features, and the regression coefficients are solved by the least squares method or gradient descent method; Finally, the mean square error and coefficient of determination are used to evaluate the model performance. The experimental results show that the top three learning behavior features that have the greatest impact on course performance are the audio and video learning time, number of chapter study times and the number of task points completed, while the multiple linear regression model established using these three learning behavior characteristics and assignment scores has the highest prediction accuracy. The research results can provide reference for online course teachers and learners, help to promote online course learning early warning and performance prediction, and improve the quality of online course teaching.