Accurate estimation of mortality in patients with cancer is important when discussing prognosis and selecting treatment. Survival estimation for many cancers is based on Tumor-Node-Metastasis (TNM) staging systems that involve three factors: tumor extent, lymph node involvement, and metastasis. The most recent clinical staging of melanoma uses TNM staging but does not include a growing number of other prognostic features. The Ensemble Algorithm of Clustering of Cancer Data (EACCD) by Chen et al. is a machine learning algorithm that regroups patients with different prognostic factors according to the survival dissimilarity. This algorithm has the potential to integrate emerging prognostic factors to more accurately stage melanoma. In this study, we use EACCD to examine the current AJCC staging of melanoma by analyzing a melanoma dataset from the National Cancer Centers Surveillance, Epidemiology, and End Rresults (SEER) database. Our results demonstrates that the EACCD algorithm generates results in-line with AJCC staging and may provide a mechanism to incorporate other prognostic factors to produce a more nuanced estimation of prognosis and survival.