Modern e-learning environments store information that enables the analysis of various aspects related to learning, their current status, and the ability to perform predictions. This paper presents a correlation analysis of students’ grades obtained from completing various assignments and tests. The analysis identifies correlations between different elements of the course, highlighting those with a high correlation to the final semester grade. The identified relationships are described using linear regression, with gradient descent employed to calculate the linear regression coefficients. The mean square error was calculated, which, on one hand, helped reduce the number of steps in calculating the linear regression coefficients, and on the other hand, allowed for an evaluation of the accuracy of the linear regression model. The analytical expression of linear regression can be used to predict students’ academic performance. Firstly, it enables students to track their academic performance and gain insight into their potential grade for the semester. Secondly, it provides teachers with an analytical tool for the early identification of underperforming students, allowing for timely motivation and the provision of educational and organizational assistance. Thus, the approach proposed in this paper, utilizing correlation and regression analysis, has significant implications for both individual students and the overall educational process.