Study on Grade Prediction and Data Analysis in Tutorial System by Deep Learning Technology Based on Modified LSTM Algorithm

This paper discusses the prediction of students' academic achievement and the data analysis under the tutorial system. In this study, we innovatively adopted the optimized and improved Long Short-Term Memory (LSTM) neural network model. This deep learning algorithm effectively solved the problem of gradient disappearance and long-term dependence existing in traditional methods when processing time series data with its unique gating mechanism. By collecting and integrating historical data on a large number of students' academic performance and participation in instructor-led activities, we input these complex time series data into an improved LSTM model for training and learning. The empirical study shows that the model can effectively capture the long-term trend and short-term fluctuation affecting the change of student achievement, and accurately predict the future academic achievement of students, and its prediction accuracy is significantly improved compared with the traditional prediction model.