Systems and methods for partitioning geographic regions
Systems, methods, and non-transitory computer-readable media can determine training data describing respective relationships between a set of map tiles, the map tiles collectively representing a given geographic region. A model can be trained to predict a likelihood of a pair of map tiles corresponding to one or more geographic classifications based at least in part on the training data. Polygons that correspond to respective sub-regions within the geographic region can be determined based at least in part on the model.
1. A computer-implemented method comprising:
determining, by a computing system, training data describing respective relationships between a set of map tiles, the map tiles collectively representing a given geographic region, wherein the training data includes a set of training examples that each includes a first map tile, a second map tile, and one or more geographic classifications that indicate whether the first map tile and the second map tile were assigned to the same territorial unit;
training, by the computing system, a machine learning model to predict a likelihood of a pair of map tiles corresponding to the one or more geographic classifications based at least in part on the training data; and
generating, by the computing system, one or more polygons that correspond to respective sub-regions within the geographic region based at least in part on the machine learning model.
2. The computer-implemented method of claim 1 , wherein a polygon corresponding to a given sub-region represents a predicted geographic classification for the sub-region.
3. The computer-implemented method of claim 1 , wherein the training examples each includes information describing one or more relationships between the first map tile and the second map tile.
4. The computer-implemented method of claim 3 , wherein a relationship between the first map tile and the second map tile is measured based on social relationships between residents of a geographic region corresponding to the first map tile and residents of a geographic region corresponding to the second map tile.
5. The computer-implemented method of claim 4 , wherein the social relationships are based on a density of social connections between users of a social networking system that reside in the geographic region corresponding to the first map tile and users that reside in the geographic region corresponding to the second map tile.
6. The computer-implemented method of claim 3 , wherein a relationship between the first map tile and the second map tile is measured based on a demographic similarity between residents of a geographic region corresponding to the first map tile and residents of a geographic region corresponding to the second map tile.
7. The computer-implemented method of claim 2 , wherein the territorial unit corresponds to a census tract, city, county, zip code, metropolis, micropolis, or any combination thereof.
8. The computer-implemented method of claim 3 , wherein a geographic classification of the first map tile and the second map tile indicates respective classifications of the first map tile and the second map tile for a set of hierarchical territorial units.
9. The computer-implemented method of claim 1 , wherein generating the one or more polygons further comprises:
clustering, by the computing system, the set of map tiles based on respective likelihoods for each pair of map tiles as predicted by the machine learning model.
10. A system comprising:
at least one processor; and
a memory storing instructions that, when executed by the at least one processor, cause the system to perform:
determining training data describing respective relationships between a set of map tiles, the map tiles collectively representing a given geographic region, wherein the training data includes a set of training examples that each includes a first map tile, a second map tile, and one or more geographic classifications that indicate whether the first map tile and the second map tile were assigned to the same territorial unit;
training a machine learning model to predict a likelihood of a pair of map tiles corresponding to the one or more geographic classifications based at least in part on the training data; and
generating one or more polygons that correspond to respective sub-regions within the geographic region based at least in part on the machine learning model.
11. The system of claim 10 , wherein a polygon corresponding to a given sub-region represents a predicted geographic classification for the sub-region.
12. The system of claim 10 , wherein the training examples each includes information describing one or more relationships between the first map tile and the second map tile.
13. The system of claim 12 , wherein a relationship between the first map tile and the second map tile is measured based on social relationships between residents of a geographic region corresponding to the first map tile and residents of a geographic region corresponding to the second map tile.
14. The system of claim 13 , wherein the social relationships are based on a density of social connections between users of a social networking system that reside in the geographic region corresponding to the first map tile and users that reside in the geographic region corresponding to the second map tile.
15. A non-transitory computer-readable storage medium including instructions that, when executed by at least one processor of a computing system, cause the computing system to perform a method comprising:
determining training data describing respective relationships between a set of map tiles, the map tiles collectively representing a given geographic region, wherein the training data includes a set of training examples that each includes a first map tile, a second map tile, and one or more geographic classifications that indicate whether the first map tile and the second map tile were assigned to the same territorial unit;
training a machine learning model to predict a likelihood of a pair of map tiles corresponding to the one or more geographic classifications based at least in part on the training data; and
generating one or more polygons that correspond to respective sub-regions within the geographic region based at least in part on the machine learning model.
16. The non-transitory computer-readable storage medium of claim 15 , wherein a polygon corresponding to a given sub-region represents a predicted geographic classification for the sub-region.
17. The non-transitory computer-readable storage medium of claim 16 , wherein the training examples that each includes information describing one or more relationships between the first map tile and the second map tile.
18. The non-transitory computer-readable storage medium of claim 17 , wherein a relationship between the first map tile and the second map tile is measured based on social relationships between residents of a geographic region corresponding to the first map tile and residents of a geographic region corresponding to the second map tile.
19. The non-transitory computer-readable storage medium of claim 18 , wherein a wherein the social relationships are based on a density of social connections between users of a social networking system that reside in the geographic region corresponding to the first map tile and users that reside in the geographic region corresponding to the second map tile.