Exploration of Optimisation Mechanism of Innovation and Entrepreneurship and Student Management for College Students Driven by Big Data

The current employment situation is getting tighter and tighter, and innovation and entrepreneurship of college students have gradually become the focus of attention of the government and colleges and universities. This paper preprocesses the student data, constructs the data matrix using the gray prediction algorithm, uses the least squares method, establishes the differential equation solution, and obtains employment prediction data. For the employment prediction can be known to the innovation and entrepreneurship situation of college students, this paper further introduces the Logistic regression analysis, establishes a multivariate regression analysis algorithm based on the influencing factors of innovation and entrepreneurship, and provides theoretical support for the study of the influencing factors and the optimization mechanism of the management of innovation and entrepreneurship student management. This paper takes University H in Zhejiang Province of China as the research site to explore the current situation of innovation and entrepreneurship of its college students in depth. The students who choose self-help entrepreneurship in 2022-2023 only accounted for 2.42% and 2.04%, and the predicted percentage in 2024 is 1.93%, so the current situation of innovation and entrepreneurship is dismal. In the analysis of college students' innovation and entrepreneurship influencing factors, the factors of specialty, personal work experience, and buddy entrepreneurship experience all showed significant differences (P<0.05). In the regression analysis of entrepreneurship education and entrepreneurship policy on entrepreneurial intention, the regression coefficients of entrepreneurship education and entrepreneurship policy are 0.276 and 0.136, respectively, which both reach the level of significance, and all of them are in a significant positive proportional relationship. © 2024 Zhengrong Ding et al., published by Sciendo.