Practical vocational education needs to lay stress on the integration of theoretical knowledge of common courses such as PE and other cultural courses with the professional knowledge of various high vocational courses. However, existing studies on practical vocational education emphasize more on theoretical research, while the practical and empirical research is slightly insufficient. For this reason, this paper attempts to analyze the feasibility of practical vocational education in higher education institutions. At first, the paper analyzed the influencing factors of the effect of practical vocational education on vocational college students, and adopted Partial Least Squares (PLS) regression to explain the independent variable and 17 dependent variables of the said effect and the relationship between each evaluation index and the corresponding evaluation criterion. Then, to verify the feasibility of practical vocational education in regional higher education institutions, this paper employed an optimized Back Propagation Neural Network (BPNN) to predict the effect of practical vocational education on vocational college students. At last, the experimental results proved the effectiveness of the constructed model and gave the conclusion of feasibility analysis.

Scientific and technological innovation ability of students is an important part of scientific and technological innovation in higher vocational colleges. There are few scientific and technological innovation achievements and low scientific and technological content of technical application achievements in higher vocational colleges at present, which are mainly caused by the single innovation organization form and weak innovation knowledge sharing ability. It is of great significance and value to explore the knowledge flow of scientific and technological innovation in colleges and universities and the collaborative innovation of college students. There is little research on the relationship between knowledge flow of scientific and technological innovation and collaborative innovation of college students in the existing literature, so this article has carried out relevant research. This article constructs the mathematical model of knowledge flow of scientific and technological innovation in colleges and universities, analyzes the stability of the mathematical model of knowledge flow of scientific and technological innovation in colleges and universities, and gives the framework of collaborative innovation ecosystem for college students. This article accurately measures the embedding of scientific and technological innovation resources in colleges and universities, the knowledge flow of scientific and technological innovation and the collaborative innovation performance of college students, and empirically tests the direct influence effect of embedding scientific and technological innovation resource network in colleges and universities on the collaborative innovation performance of college students and the moderating effect of the knowledge flow of scientific and technological innovation in colleges and universities. Finally, the corresponding experimental analysis results are given.

With the shift of the global education view to humanism, more and more scholars and experts have gradually deepened their research on sustainable development ability of students in higher vocational colleges. As for sustainable development of students, the core and basic point is to develop their personality and teach them in accordance with their aptitude. As a bridge and link between teaching thinking and practice of teachers, their teaching decisions need to fully take into consideration the sustainable development of students, which is of great significance to future employment, job transfer and promotion of students. Therefore, this paper studies the teaching decision optimization (TDO) for the sustainable development of students. First of all, a related teaching decision model is constructed. Then this paper proposes the particle swarm optimization algorithm (PSO) of teaching decision path based on progress rate, gives the principle of selecting TDO template, designs three different optimization models for ordinary particles, high-quality particles and progress particles and represents the optimization strategy of teaching decision path. Finally, experimental results verify the effectiveness of the optimization model.