The construction of modern aesthetic education innovative talent cultivation mode in universities based on big data technology

Abstract To strengthen the comprehensive literacy of college students and have a correct aesthetic concept, cultivation innovative talents in modern education under big data is proposed. The mixed-filtered training were used to calculate prediction function, loss regression parameters using logistic model. sample probabilities are derived from maximum likelihood function predict unmade scores teacher’s scoring matrix, which trained fit original matrix. behavior each student divided into corresponding label, courses with historical preferences same category recommended target that category. By iteratively updating values cluster center one by one, high similarity clustered class, distances all centers recalculated alleviate sparsity problem scalability regression. information set as input data, numerical type indicators basis for judgment obtain satisfaction situation results, visually evaluating gap between predicted real ratings system accurate complete on university counted. analysis results show technology, physical colleges universities carry out concept beauty, form better, percentage strength beauty reaches 69%, strengthening ability feel beauty.