Predicting course difficulty based on grades in supporting disciplines using logistic regression: a case study of a Python programming course

Бесплатный доступ

The article discusses methods for predicting the difficulty of academic courses based on logistic regression using grades from prerequisite subjects. The main subject of the study is the course "Programming in Python" for which key prerequisite subjects are mathematics, computer science, and English. The aim of the study is to develop a model that allows for the adaptation of academic assignments to the individual needs of students, thereby enhancing the effectiveness of the educational process. Synthetic data is used to implement the model due to limitations in access to real educational data. The application of machine learning methods, particularly logistic regression, allows not only for the classification of courses by difficulty (easy, medium, hard) but also for probabilistic assessments that reflect the model's confidence in its predictions. The authors examine the weight coefficients of the features, which allows for an understanding of the contribution of each prerequisite subject to the prediction of difficulty. Predicting the difficulty of courses and assignments facilitates more accurate selection of educational materials, improving the quality of education and promoting the development of personalized educational trajectories. Thus, the article contributes to the advancement of educational analytics methods and emphasizes the need to transition from predicting student performance to predicting course difficulty, opening up new prospects for the personalization and enhancement of the educational process.

Еще

Individual learning paths, course difficulty, supporting disciplines, machine learning, edm, learning analytics, assignment selection, logistic regression

Короткий адрес: https://sciup.org/14131166

IDR: 14131166

Статья научная