Analysis of the results of adaptive learning of students using LSTM neural network

The object of this study is the process of adaptive learning management using artificial intelligence methods. The paper explores the application of neural networks for analyzing student responses and identifying knowledge gaps. The primary focus is on evaluating the effectiveness of the Long Short-Term Memory (LSTM) model in predicting test results based on a student's response history. The proposed approach is based on analyzing the sequence of student responses and utilizing training data to detect patterns in academic performance. A comparison with other machine learning means, such as gradient boosting and random forest, demonstrated the competitive accuracy of LSTM. The most effective approach not only predicts the correctness of responses but also identifies potential difficulties in understanding the material. The study results indicate that the use of neural network models in the educational process enhances the efficiency of adaptive learning. Analyzing key learning patterns has allowed for the identification of major factors influencing student performance. It was found that LSTM can adjust individual learning trajectories by recommending additional materials or modifying the teaching approach for topics that present difficulties. The proposed methodology can be used in the development of automated teacher support systems and personalized learning environments. It can be integrated into digital educational platforms to detect weak points in students' knowledge and adapt the learning process to their individual needs. The integration of such technologies is expected to improve knowledge retention and reduce the frequency of repeated mistakes during learning.