Study on Score Prediction Model with High Efficiency Based on Deep Learning

Abstract

In the problem of unified classroom performance prediction, there is a certain lag in the prediction, and there are also problems such as the data sparsity and single feature in the data. In addition, feature engineering is often carried out manually in modeling, which highly depends on the professional knowledge and experience of engineers and affects the accuracy of the prediction to a certain extent. To solve the abovementioned gaps, we proposed an online course score prediction model with a high time efficiency that combines multiple features. The model uses a deep neural network, which can automatically carry out feature engineering and reduce the intervention of artificial feature engineering, thus significantly improving the time efficiency. Secondly, the model uses a factorization machine and two kinds of neural networks to consider the influence of first-order features, second-order features, and higher-order features at the same time, and it fully learns the relationship between the features and scores, which improves the prediction effect of the model compared to using only single feature learning. The performance of the model is evaluated on the learning analysis dataset from Fall 2015 to Spring 2021 and includes 412 courses with 600 students. The experimental results show that the performance of the prediction model based on the feature combination proposed in the present study is better than the previous performance prediction model. More importantly, our model has the best time efficiency of below 0.3 compared to the other models.

Deep Learning Techniques for Predicting Student’s Academic Performance on Virtual Learning Environments: A Review

“…The best strategy to enhance the DNN's performance is to experiment with various hyperparameters. Yang and Bai (2022). As shown in Figure 4, using DNN architecture in predictive models has produced a high accuracy of above 70% and up to 97.5%, and on average 85.89%, except for Ángel Casado Hidalgo et al (2021) which has achieved about 67.2%, that because the model has used a limited dataset from LMS whereas the DL predictive models need for a massive dataset to perform well.…”

“…Table 10 describes the datasets that are used in predicting student's academic performance including the number of students and courses used with each one of them. Waheed et al (2019Waheed et al ( , 2022; ; Yang and Bai (2022).…”

“…A self-attention mechanism and depth matrix decomposition were integrated into the student performance prediction model using DNNs, the model suggested by Yang and Bai (2022) is called (Factorization Deep Product Neural network, FDPN). The FDPN model fully learns the relationship between the features and scores, which improves the prediction effect of the model in comparison with using only single feature learning.…”

“…Similarly, the study by Yang and Bai (2022), proposed a Factorization Deep Product Neural network model (FDPN) using a sports education dataset available at OULAD for course score prediction. The study discovered that using more than two features simultaneously had some impact on the results of student-grade prediction.…”

“…This reflects their knowledge of the concepts taught in these courses and the skills they cover. Typically, a range of evaluation techniques, such as exams, tests, essays, projects, and other assignments, are used to determine the overall academic performance and achievement Aljaloud et al (2022); Yang and Bai (2022). Overall academic outcomes may consider additional aspects of academic success, such as behavior, attendance, and social-emotional development.…”