Abstract: The fresh agricultural product supply chain is a complex system network that covers lots of links from production to circulation and to consumption. Efficient production and management are of great significance for ensuring the quality and safety of fresh produce. As one of the major fields of information science, machine learning has gradually achieved wide application and effective results in many aspects of the agricultural products supply chain. This paper selects five key links of the fresh agricultural product supply chain and comprehensively summarizes the application research status and used methods of machine learning, especially deep learning technology in these fields in the past five years. Finally, the data and algorithm types used in mainstream research are summarized, and the future research trends are prospected. It is pointed out that future research can be carried out from three aspects: the establishment of standard datasets, the construction of models based on prior knowledge, and the promotion of data integration and sharing.