Dry port is an inland intermodal terminal directly linked by road or rail to a seaport as a center for the transshipment of sea freight to inland destinations. However roads contribute to pollution more than rail. For sustainable transportation systems, we need to improve the performance of rail dry port systems. The purpose of this research is to propose a diagnostic system for modeling and classification of rail dry port performance, by use of supervised ensemble learning approaches. We propose an empirical comparison of supervised ensemble learning approaches and application for performance diagnosis of dry port by rail. Our results demonstrate that the performance testing of proposed diagnostic system is found to be very satisfactory. The calculation methods demonstrating the results of the data are provided in this study diagnosis of dry port by rail. Finally, the conclusions for classification modeling for dry port by State Railway of Thailand (SRT) are given.