Systems for machine-learned resource availability determination for a population

1 . A computer-implemented method for determining a social health score for a region, the computer-implemented method comprising:

receiving, at processor hardware, first data for a geographical area, wherein the first data includes values for a plurality of features, and wherein the plurality of features corresponds to a plurality of domains;

for a first domain of the plurality of domains:

from the plurality of features, selecting a first set of features that correspond to the first domain, and

determining, via the processor hardware, first feature weights for the first set of features based on a feature-level quantile g-computation of the values of the first data for the first set of features, wherein the first feature weights indicate relative amounts of variation contributed by the respective features to the first domain;

receiving, at the processor hardware, second data for the region of the geographical area, wherein the second data includes values for a subset of the plurality of features;

determining, via the processor hardware, whether the first domain is a first domain type;

in response to a determination that the first domain is the first domain type, determining, via the processor hardware, a mean value for a respective feature of the subset of the plurality of features by:

identifying an area of an image in the second data corresponding to the region,

determining a pixel value average of the area of the image, and

assigning the pixel value average as the mean value;

for the first domain, determining, via the processor hardware, a first region domain score based on the second data and the first feature weights;

determining, via the processor hardware, a total score for the region based on the first region domain score, wherein the total score indicates a level of resource access for a population of the region; and

transforming a user interface to reflect the total score.

2 . The computer-implemented method of claim 1 wherein the total score indicates a level of impediment for the population of the region of the geographical area to access resources.

3 . The computer-implemented method of claim 1 further comprising:

determining feature scores for the subset of the plurality of features by cross-multiplying the first feature weights and the values of the second data, and

determining the first region domain score based on the feature scores.

4 . The computer-implemented method of claim 3 wherein determining the first region domain score includes aggregating the feature scores.

5 . The computer-implemented method of claim 1 further comprising:

for a second domain of the plurality of domains:

from the plurality of features, selecting a second set of features that correspond to the second domain, and

determining second feature weights for the second set of features based on the feature-level quantile g-computation of the values of the first data for the second set of features, wherein the second feature weights indicate relative amounts of variation contributed by the respective features to the second domain; and

for the second domain, determining a second region domain score based on the second data and the second feature weights,

wherein the total score of the region is based on aggregating the first region domain score and the second region domain score.

6 . The computer-implemented method of claim 1 wherein there is a one-to-one correspondence between the first feature weights and the first set of features.

7 . The computer-implemented method of claim 1 further comprising in response to the subset being a proper subset, fill in those missing features of the second data.

8 . The computer-implemented method of claim 1 wherein the region is a United States census tract.

9 . The computer-implemented method of claim 1 further comprising:

determining a domain score for the first domain based on the first feature weights and the first data;

determining a domain weight for the first domain based on a domain-level quantile g-computation applied to the domain score, wherein the domain weight indicates a relative variation the first domain contributes to the geographical area; and

determining the first region domain score based on the domain weight.

10 . The computer-implemented method of claim 9 further comprising:

determining feature scores based on cross-multiplying the first feature weights and the values of the second data; and

determining the first region domain score based on cross-multiplying the domain weight and the feature scores.

11 . The computer-implemented method of claim 1 wherein the plurality of domains includes seven domains, and the plurality of features includes twenty-eight features.

12 . The computer-implemented method of claim 1 further comprising preprocessing the values for the plurality of features to fit within a range.

13 . The computer-implemented method of claim 12 wherein fitting the values within the range includes changing at least one value that is lower than the range to a minimum value of the range.

14 . The computer-implemented method of claim 12 wherein fitting the values within the range includes changing at least one value that is higher than the range to a maximum value of the range.

15 . The computer-implemented method of claim 1 wherein none of the first feature weights are below five percent.

16 . The computer-implemented method of claim 1 wherein:

the first data includes:

a plurality of life expectancy values for a plurality of regions of the geographical area,

a plurality of rural status values for the plurality of regions, and

a plurality of census region values for the plurality of regions; and

the first feature weights are determined based on the plurality of life expectancy values, the plurality of rural status values, and the plurality of census region values.

17 . A system for determining a scoring a region, the system comprising:

a database including:

first data for a geographical area, wherein the first data includes values for a plurality of features, and wherein the plurality of features corresponds to a plurality of domains, and

second data for the region of the geographical area, wherein the second data includes values for a subset of the plurality of features;

a weighting module configured to, for a first domain of the plurality of domains:

from the plurality of features, select a first set of features that correspond to the first domain, and

determine, via processor hardware, first feature weights for the first set of features based on a feature-level quantile g-computation of the values of the first data for the first set of features, wherein the first feature weights indicate relative amounts of variation contributed by the respective features to the first domain;

a domain scoring module configured to;

determine, via the processor hardware, whether the first domain is a first domain type;

in response to a determination that the first domain is the first domain type, determine, via the processor hardware, a mean value for a respective feature of the subset of the plurality of features by:

identifying an area of an image in the second data corresponding to the region,

determining a pixel value average of the area of the image, and

assigning the pixel value average as the mean value; and

determine, via the processor hardware, for the first domain, a first region domain score based on the second data and the first feature weights;

a region scoring module configured to determine, via the processor hardware, a total score for the region based on the first region domain score, wherein the total score indicates a level of resource access for a population of the region; and

a user interface to reflect the total score.

18 . The system of claim 17 wherein the domain scoring module is configured to:

determine feature scores for the subset of the plurality of features by cross-multiplying the first feature weights and the values of the second data, and

determining the first region domain score based on the feature scores.

19 . The system of claim 17 wherein:

the weighting module is configured to, for a second domain of the plurality of domains:

from the plurality of features, select a second set of features that correspond to the second domain, and

determine second feature weights for the second set of features based on the feature-level quantile g-computation of the values of the first data for the second set of features, wherein the second feature weights indicate relative amounts of variation contributed by the respective features to the second domain;

the domain scoring module is configured to, for the second domain, determine a second region domain score based on the second data and the second feature weights; and

the region scoring module is configured to determine the total score of the region based on aggregating the first region domain score and the second region domain score.

20 . A non-transitory computer-readable medium comprising instructions including:

receiving, at processor hardware, first data for a geographical area, wherein the first data includes values for a plurality of features, and wherein the plurality of features corresponds to a plurality of domains;

for a first domain of the plurality of domains:

from the plurality of features, selecting a first set of features that correspond to the first domain, and

determining, via the processor hardware, first feature weights for the first set of features based on a feature-level quantile g-computation of the values of the first data for the first set of features, wherein the first feature weights indicate relative amounts of variation contributed by the respective features to the first domain;

receiving, at the processor hardware, second data for a region of the geographical area, wherein the second data includes values for a subset of the plurality of features;

determining, via the processor hardware, whether the first domain is a first domain type;

in response to a determination that the first domain is the first domain type, determining, via the processor hardware, a mean value for a respective feature of the subset of the plurality of features by:

identifying an area of an image in the second data corresponding to the region,

determining a pixel value average of the area of the image, and

assigning the pixel value average as the mean value;

for the first domain, determining, via the processor hardware, a first region domain score based on the second data and the first feature weights;

determining, via the processor hardware, a total score for the region based on the first region domain score, wherein the total score indicates a level of resource access for a population of the region; and

transforming a user interface to reflect the total score.