A Score Prediction Model based on Deep Learning for Teaching Quality Analysis

Kan Ji, Bangying Wu, Qiang Yao

Abstract: Based on the increasing amount of campus data, more and more scholars are starting to explore potential patterns from it to improve teaching mode and promote teaching quality. Teaching quality analysis has gradually become a hot research area. The-state-arts-of approaches such as big data analysis, data mining, machine learning, and deep learning have begun to be widely applied in research on teaching quality, and student score is one of the crucial indicators for measuring teaching quality. In this paper, we construct a score prediction model based on deep learning to analyze the teaching quality of teacher. This model can also be applied to evaluation indicators for teaching quality analysis other than student score. The score prediction model based on neural networks solves the problem of inaccurate evaluation results of traditional teaching quality evaluation methods on nonlinear regression models. Based on known student score, a score prediction model based on deep learning can help teacher assess student' mastery level of course knowledge precisely and predict student final exam score to achieve the effect of early warning for student in the final exam.

Keywords: Teaching quality analysis, deep learning, neural networks, student score, early warning