【Objective】The spatial distribution of mineralogy classes in mineral soils reflects the regional pedogenic environment and soil-forming processes, and this information is essential for understanding soil genesis and development.【Method】This study focuses on Hubei Province, a region with a subtropical climate in China. By integrating legacy soil mineralogy class data, a fuzzy mapping framework was proposed based on an expert knowledge decision tree. The study further investigated the spatial probability distribution and driving factors of soil mineralogy classes using three different models. 【Result】The results showed that the CatBoost model demonstrated the best performance among the three models. The soil mineralogy classes in the study area were predominantly Mixed, Illite mixed, and Siliceous mixed, while Smectite, Kaolinite, and Vermiculite classes were distributed in scattered patches. Based on the validation results, there were differences in mapping accuracy among different mineralogy classes. Among these, mapping results for mixed types, such as Siliceous mixed, were relatively good, while those for single types exhibited higher uncertainty. Further investigation of the characteristics of environmental variables revealed that hydrothermal conditions may influence the spatial differentiation of Mixed and Kaolinite mixed by affecting the intensity of weathering and leaching, while the parent material makes a significant contribution to the spatial distribution of Siliceous mixed and the Smectite mixed.【Conclusion】These research findings contribute to a deeper understanding of soil formation and evolution processes in the subtropical transition zone and provide a reference basis for the systematic classification and naming of soil families in the 3rd National Soil Census.