Quality Evaluation of College Students' Innovation and Entrepreneurship Education Based on the K-Means Clustering Algorithm

Abstract

Innovation and entrepreneurship are increasingly recognized as vital components of modern higher education, fostering creativity, problem-solving skills, and economic growth. However, effectively catering to the diverse needs of students in these domains remains a challenge. This paper proposes a novel approach to address this challenge by applying the Cluster Swap Expectation Maximization (CSSEM) algorithm to cluster college students based on attributes relevant to innovation and entrepreneurship. Using a dataset encompassing metrics such as GPA, innovation score, entrepreneurship score, socioeconomic status, and gender, we demonstrate the effectiveness of the CSSEM algorithm in segmenting student populations. The findings reveal distinct clusters of students with varying levels of academic performance, innovation potential, and entrepreneurial aspirations. For instance, Cluster 1 comprises students with lower GPAs (mean GPA = 3.2) and moderate innovation and entrepreneurship scores (mean innovation score = 78, mean entrepreneurship score = 80), while Cluster 2 consists of high-performing students (mean GPA = 3.7) with strong innovation and entrepreneurship potential (mean innovation score = 90, mean entrepreneurship score = 88). These insights enable educators and policymakers to design tailored interventions and support mechanisms that cater to the specific needs of different student groups.