Predictive Modeling Using Logistic Regression on New Media Data for College Student Entrepreneurship

With the rapid development of new media technology, returning to their hometown for employment has gradually become an important choice for college students, alleviating the employment pressure in big cities and injecting new impetus into regional economic development. This paper takes 287 college students from three universities in the northern district of Zhengzhou as the survey objects, collects data through a questionnaire survey and uses descriptive statistical analysis, chi-square test and logistic regression analysis to explore the current situation and influencing factors of college students returning to their hometown for employment under the background of new media. The study found that: first, college students have a low willingness to return to their hometown for employment, which is mainly affected by insufficient infrastructure in their hometown, low policy awareness, and unclear career planning; second, college students prefer stable jobs and have a weak willingness to start a business; third, satisfaction with hometown development, family support and the practice of “three going to the countryside” are key influencing factors.