 This paper proposes a deep learning-based stacked autoencoder model to classify the operating state of a ship based on its power load data. The model was evaluated against real data and compared to other models. The results showed that the proposed model had a higher true positive rate, lower false positive rate, higher Matthews correlation coefficient, and higher accuracy than other models. This indicates that deep learning can be used to accurately classify ship operating modes. This article was authored by Ji-Yoon Kim and Jeanne Suk-Oh.