We applied an unsupervised clustering algorithm, initially trained on data from Magnetospheric Multiscale (MMS) mission at Earth, to MESSENGER observations at Mercury to identify three distinct plasma regions: magnetosphere, magnetosheath, and solar wind. This demonstrates the applicability of transfer learning for heliophysics, a machine learning technique where knowledge learned from one task is reused to perform a similar task, for an unsupervised learning task. We find that 72% of bow shock crossings and 84% of magnetopause crossings identified by the clustering algorithm using MESSENGER data are in agreement with published magnetopause and bow shock lists. These findings highlight the potential for utilizing a clustering algorithm across multiple planetary environments.