Traditional models for analyzing learning ability use simple, single-factor models that cannot be analyzed to impact analyze the key factors that affect learning. Therefore, the study establishes a learning ability optimization model for higher vocational students based on principal component analysis combined with back propagation neural network. By collecting multi-source heterogeneous data such as class activity and discussion activity, and using back propagation neural network to provide students with corresponding optimization of learning ability. The experiment builds its own database of multiple students from higher vocational schools, and the research model has a mean absolute error of 0.23 points in predicting scores when the number of iterations reaches 400, and the root mean square error stabilizes at 0.66 points at the end of the validation set of iterations. In the comparison of the optimization of the students' scores it was shown that the research model improved the score of student #46 by 12 points. The above results show that the research model can target and analyze the learning ability of higher vocational students one by one, and provide new reference ideas for creating a good academic style in higher vocational colleges. © 2025 IEEE.
Author Keywords
Index Keywords
