Year: 2026
Type: conference-paper
Abstract: Student achievement prediction can effectively provide data support for educational reform and educational decision-making for students and parents. Traditional performance prediction methods often focus on single factors such as exam scores, without considering social factors such as educational background, study habits, and family environment. In order to reduce prediction of student performance prediction, a performance prediction method by multi feature fusion is proposed, fully considering the influence of social factors. Classification predictions were made based on the differences in characteristics among different groups. The feature selection strategy based on T-test and sequential forward search not only reduces the dimensionality of feature data, but also preserves useful information and improves the computational efficiency of the achievement prediction. By analyzing data from different regions, the accuracy and computational efficiency of the algorithm have been verified.
Source: Procedia Computer Science
Authors: Na Wang, Jinsong Li, Hongyan Liu, Minghai Yao
Institution: Bohai University
Language: English
Subfield: Computer Science Applications
Field: Computer Science
Domain: Physical Sciences
SDG: Quality education
Open Access status: diamond
OpenAlex ID: W7203836631