Venture Capital Project Segmentation Supporting College Students’ Entrepreneurial Decisions

College students innovative undertaking can help improve the vitality of regional economic development. However, the success rate of college students’ entrepreneurship is not high due to the problems of inexperience and lack of capital. Some efforts are needed to ensure successful entrepreneurship for college students. Existing studies mainly focus on entrepreneurs themselves and the impact of the external environment on entrepreneurship without considering the abundant information in venture capital data. This study mines the data of venture capital projects from the perspective of entrepreneurial projects and investment events. Clustering, as a typical data mining task, can induce the commonness of similar data and distinguish the individuality of different data. Besides, clustering method is more suitable for handling unlabeled venture capital data than other supervised methods such as classification. Therefore, this study establishes a model of venture capital project segmentation based on clustering techniques. Specifically, to cope with the characteristics that the collected entrepreneurial data contain different types of attributes, the clustering algorithm for incomplete data sets with mixed numeric and categorical attributes is adopted and improved to subdivide entrepreneurial projects and investment events into five types, respectively. By analyzing the characteristics of each type, we summarize the general pattern of the same type of entrepreneurial projects and investment events to support college students’ entrepreneurial decisions or school and social assistance policy formulation.