Constructing an innovative system of management and education in colleges and universities based on artificial intelligence technology

Abstract

At this stage, the information literacy of teachers and students is constantly improving, and the “digital campus” represented by artificial intelligence is landing in various regions and schools. This paper accurately constructs a college management and education innovation system based on artificial intelligence technology, including institutional mechanisms, a one-stop service platform, and an early warning mechanism. The early warning mechanism has an early warning model constructed. The construction of an academic early warning model for students is based on multivariate linear regression. A model for early warning of student activity safety was constructed using multiclassification logistic regression. The impact of fusing the two models was evaluated, as well as the impact of practical application. The ROC value of the first-level warning is 0.9533, and the ROC value of the second-level warning is 0.9428, and both values are close to 1, which verifies that the model can be practically applied. By analyzing the F1-score value through practical application, the average value of the F1-score throughout the semester was as high as 84.68%, and the number of warnings of the early warning model was higher than the number of feedback from teachers. It has been verified that the early warning model in this study can effectively and efficiently manage student activities and academic anomalies in colleges and universities.

Keywords

• multiple linear regression
• logistic regression
• early warning model
• college management
• 68t05