Innovation of educational management paths in higher education based on LSTM deep learning model

Abstract

Abstract In this paper, based on the data of the Riiid education platform, the LSTM deep learning model is used to provide accurate prediction and guidance for the education management of colleges and universities. The Gini coefficient is also introduced to simplify the calculation process, focusing on predicting the development of students’ careers. To achieve this goal, the online education platform provided a dataset that was carefully pre-processed and cleaned of data, and feature engineering was performed to obtain more informative features. Comparing the AUC value of the offline area of the ROC curve, the AUC value of the LSTM deep learning model can reach 0.758, and the training time of a single model is about 41.8 seconds. Therefore, a deep learning model based on the LSTM algorithm can be used for innovation research.