Student entrepreneurship plays a vital role in driving regional economic vitality and innovation. Accurately identifying its key determinants is essential for developing effective support policies. This study targets university students from selected colleges in Zhejiang Province, collecting data on their entrepreneurial activities through structured questionnaires. To improve data analysis reliability, an enhanced Semi-Supervised Fuzzy C-Means (SS-FCM) clustering model is employed to preprocess the raw data and categorize features. By integrating limited labeled data with a large volume of unlabeled data, the SS-FCM model effectively captures fuzzy boundaries and hidden patterns, thereby enhancing the accuracy of variable selection and model construction. Based on the clustering results, key factors influencing entrepreneurial success are identified, and an index-based evaluation system is established to quantitatively assess their impact. Finally, targeted support strategies and policy recommendations are proposed to improve the success rate and sustainability of student entrepreneurship. The findings offer valuable insights for policymakers, university administrators, and student entrepreneurs, providing practical guidance and strategic direction.