Disaggregating Census Data for Population Mapping Using Random Forests with Remotely-Sensed and Ancillary Data
Fig 1
This figure represents the general structure of the data processing and map production procedure used to compare the methodology outlined in this paper to the AfriPop/AsiaPop, GRUMP, and GPW methodologies.
The orange boxes represent items that are specific to the research presented here and not part of end-user map data product generation. The green boxes represent data pre-processing stages. Items in blue represent Random Forest model estimation, per-pixel prediction and dasymetric redistribution of census counts.