In this paper, an association matrix is constructed based on a global self-representation model, and the problem of constructing an association matrix with sparse self-representation of samples and Laplace regularized least squares classifier are integrated to improve the model performance. Based on the block coordinate descent algorithm, iterative solving for each variable was performed to optimize the model, and then the decision-making model was constructed using the multi-objective decision-making algorithm in order to provide managers with appropriate educational and management program services. Finally, the influence of ideology and political education on students’ quality as well as college management is analyzed, and the way of correlation evolution between ideology and politics and student management is explored. The results show that ideology and politics have a positive influence on student quality, and the degree of influence on personal concepts as well as on the pursuit of knowledge and the establishment of goals is 0.75 and 0.8, respectively. The probability that a highly correlated and coordinated college or university with ideology politics and student management will remain in this rank in the following year is 65%, and the probability of downward or upward development is 20% in both cases.