Educational institutions face challenges in attracting prospective students while maintaining academic quality and resource efficiency. This study applies a hybrid approach that integrates Self-Organizing Maps (SOM) and K-Means to cluster schools based on four attributes, namely the number of accounts, average UTBK scores, geographical distance, and parental income. The analysis's findings produce three distinct clusters. With a high degree of attribute variation, Cluster 2 (279 schools) is a dominant group that suggests the possibility of extensive marketing campaigns. Clusters 1 (45 schools) and 3 (81 schools), on the other hand, are more uniform and call for a more specialized and focused strategy. These results imply that a data-driven approach can help institutions create interventions that are specific to each segment's profile and increase the efficacy of educational marketing strategies. In order to improve segmentation accuracy in the future, this study creates opportunities for investigating new features and dynamic clustering techniques.
Alifah et al. (Thu,) studied this question.