A unbiased recursive partitioning-based framework including CIT and RF is developed.
The drivers of land use change at a regional scale are identified and apportioned.
Models show strong suitability in assessing the drivers of land use change.
Socioeconomic factors contribute the most to the land use change in our case.