Granted Patent B1

US 12,271,387 · App. 18/798,168 · Granted Apr 8, 2025

Inventors: Phillip H. Rogers (San Francisco, CA); Jonathan B. Ward (San Francisco, CA); Rashmi Poudel (San Francisco, CA); Emily Barry (San Francisco, CA); Melinda Sue Gomez Tellez (San Francisco, CA); Prajwal Vijendra (San Francisco, CA); Azriel S. Ghadooshahy (San Francisco, CA); Emmet Sun (San Francisco, CA)

Assignee: Clarify Health Solutions, Inc.

G06F16/24578G06F16/2282

Abstract

Claims (75)

1. A system, comprising:

one or more memory units comprising one or more instructions;

one or more processors communicatively coupled to the one or more memory units, the one or more processors configured, upon executing the one or more instructions, to:

for each of one or more lookup tables:

access a table that comprises a first set of data, wherein the first set of data comprises a plurality of first independent variables, wherein all of the plurality of the first independent variables are one-hot encoded;

rank the plurality of first independent variables, and sort the plurality of first independent variables into a first order based on the rank;

distribute the sorted first independent variables of the second table into a plurality of first bands of first varying resolution;

generate a first hash for each record of the first set of data at each level of the plurality of first bands;

generate a respective lookup table of the one or more lookup tables based on the first hashes;

access a table that comprises a second set of data, wherein the second set of data comprises a plurality of second independent variables, wherein all of the plurality of the second independent variables are one-hot encoded;

rank and sort the plurality of second independent variables into a second order that is the same as the first order of the sorted first independent variables;

distribute the sorted second independent variables into a plurality of second bands of a second varying resolution, wherein the distribution of the sorted second independent variables into the plurality of second bands of the second varying resolution is the same as the distribution of the sorted first independent variables into the plurality of first bands of the first varying resolution, wherein the plurality of second bands of the second varying resolution is the same as the plurality of first bands of the first varying resolution;

generate a second hash for each record of the second set of data at each level of the plurality of second bands, so as to create a model-ready table; and

join the respective lookup table to the model-ready table on matching hashes of the first and second hashes.

2. The system of claim 1 , wherein the one or more processors are further configured, upon executing the one or more instructions, to, for each of the one or more lookup tables, remove one or more first bands of the plurality of first bands and/or one or more second bands of the plurality of second bands based on a minimum match volume hyperparameter.

3. The system of claim 1 , wherein, for each of the one or more lookup tables, the respective lookup table comprises one or more predicted values for each of the first hashes.

4. The system of claim 3 , wherein the one or more processors are further configured, upon executing the one or more instructions, to:

receive a user prediction request from a user device;

generate a prediction report based on at least one lookup table of the one or more lookup tables; and

transmit the prediction report to the user device.

5. The system of claim 4 , wherein the user device is a member of a group comprising:

a computer;

a desktop PC;

a laptop PC;

a smart phone;

a tablet computer; and

a personal wearable computing device.

6. The system of claim 1 , wherein the one or more processors are further configured, upon executing the one or more instructions, to:

for each of the one or more lookup tables:

one-hot encode one or more of the first independent variables of the plurality of first independent variables of the first set of data, wherein the one-hot encoding results in all of the plurality of the first independent variables being one-hot encoded; and

one-hot encode one or more of the second independent variables of the plurality of second independent variables of the second set of data based on a mapping of the one-hot encoding of all of the plurality of first independent variables of the first set of data, wherein the one-hot encoding results in all of the plurality of the second independent variables being one-hot encoded.

7. The system of claim 1 , wherein, for each of the one or more lookup tables, each of the plurality of first independent variables is ranked based on a correlation between the respective first independent variable and a target metric.

8. The system of claim 1 , wherein, for each of the one or more lookup tables, the plurality of first bands comprise a top band that comprises a smallest portion of the sorted first independent variables in comparison to each of the other first bands of the plurality of first bands, wherein each of the other first bands of the plurality of first bands includes a larger portion of the sorted first independent variables in comparison to a previous first band of the plurality of first bands.

9. The system of claim 1 , wherein, for each of the one or more lookup tables, the plurality of first independent variables comprises one or more one-hot encoded categorical variables and one or more one-hot encoded numerical variables.

10. The system of claim 1 , wherein, for each of the one or more lookup tables, the respective lookup table of the one or more lookup tables is generated using one or more query language statements.

11. A method, comprising:

for each of one or more lookup tables:

accessing, by one or more processors of a computing system, a table that comprises a first set of data, wherein the first set of data comprises a plurality of first independent variables, wherein all of the plurality of the first independent variables are one-hot encoded;

ranking, by the one or more processors of the computing system, the plurality of first independent variables, and sorting, by the one or more processors of the computing system, the plurality of first independent variables into a first order based on the rank;

distributing, by the one or more processors of the computing system, the sorted first independent variables of the second table into a plurality of first bands of first varying resolution;

generating, by the one or more processors of the computing system, a first hash for each record of the first set of data at each level of the plurality of first bands;

generating, by the one or more processors of the computing system, a respective lookup table of the one or more lookup tables based on the first hashes;

accessing, by the one or more processors of the computing system, a table that comprises a second set of data, wherein the second set of data comprises a plurality of second independent variables, wherein all of the plurality of the second independent variables are one-hot encoded;

ranking and sorting, by the one or more processors of the computing system, the plurality of second independent variables into a second order that is the same as the first order of the sorted first independent variables;

distributing, by the one or more processors of the computing system, the sorted second independent variables into a plurality of second bands of a second varying resolution, wherein the distribution of the sorted second independent variables into the plurality of second bands of the second varying resolution is the same as the distribution of the sorted first independent variables into the plurality of first bands of the first varying resolution, wherein the plurality of second bands of the second varying resolution is the same as the plurality of first bands of the first varying resolution;

generating, by the one or more processors of the computing system, a second hash for each record of the second set of data at each level of the plurality of second bands, so as to create a model-ready table; and

joining, by the one or more processors of the computing system, the respective lookup table to the model-ready table on matching hashes of the first and second hashes.

12. The method of claim 11 , further comprising, for each of the one or more lookup tables, removing, by the one or more processors of the computing system, one or more first bands of the plurality of first bands and/or one or more second bands of the plurality of second bands based on a minimum match volume hyperparameter.

13. The method of claim 11 , wherein, for each of the one or more lookup tables, the respective lookup table comprises one or more predicted values for each of the first hashes.

14. The method of claim 13 , further comprising:

receiving, by the one or more processors of the computing system, a user prediction request from a user device;

generating, by the one or more processors of the computing system, a prediction report based on at least one lookup table of the one or more lookup tables; and

transmitting, by the one or more processors of the computing system, the prediction report to the user device.

15. The method of claim 11 , further comprising, for each of the one or more lookup tables:

one-hot encoding, by the one or more processors of the computing system, one or more of the first independent variables of the plurality of first independent variables of the first set of data, wherein the one-hot encoding results in all of the plurality of the first independent variables being one-hot encoded; and

one-hot encoding, by the one or more processors of the computing system, one or more of the second independent variables of the plurality of second independent variables of the second set of data based on a mapping of the one-hot encoding of all of the plurality of first independent variables of the first set of data, wherein the one-hot encoding results in all of the plurality of the second independent variables being one-hot encoded.

16. The method of claim 11 , wherein, for each of the one or more lookup tables, the ranking, by the one or more processors of the computing system, the plurality of first independent variables comprises ranking each of the plurality of first independent variables based on a correlation between the respective first independent variable and a target metric.

17. The method of claim 11 , wherein, for each of the one or more lookup tables, the generating, by the one or more processors of the computing system, the respective lookup table of the one or more lookup tables comprises generating, by the one or more processors of the computing system, the respective lookup table of the one or more lookup tables using one or more query language statements.

18. Tangible, non-transitory computer-readable media comprising program instructions that are configured, when executed, to cause a computing system to perform functions comprising:

for each of one or more lookup tables:

accessing a table that comprises a first set of data, wherein the first set of data comprises a plurality of first independent variables, wherein all of the plurality of the first independent variables are one-hot encoded;

ranking the plurality of first independent variables, and sorting the plurality of first independent variables into a first order based on the rank;

distributing the sorted first independent variables of the second table into a plurality of first bands of first varying resolution;

generating a first hash for each record of the first set of data at each level of the plurality of first bands;

generating a respective lookup table of the one or more lookup tables based on the first hashes;

accessing a table that comprises a second set of data, wherein the second set of data comprises a plurality of second independent variables, wherein all of the plurality of the second independent variables are one-hot encoded;

ranking and sorting the plurality of second independent variables into a second order that is the same as the first order of the sorted first independent variables;

distributing the sorted second independent variables into a plurality of second bands of a second varying resolution, wherein the distribution of the sorted second independent variables into the plurality of second bands of the second varying resolution is the same as the distribution of the sorted first independent variables into the plurality of first bands of the first varying resolution, wherein the plurality of second bands of the second varying resolution is the same as the plurality of first bands of the first varying resolution;

generating a second hash for each record of the second set of data at each level of the plurality of second bands, so as to create a model-ready table; and

joining the respective lookup table to the model-ready table on matching hashes of the first and second hashes.

19. The tangible, non-transitory computer-readable media of claim 18 , wherein, for each of the one or more lookup tables, the respective lookup table comprises one or more predicted values for each of the first hashes.

20. The tangible, non-transitory computer-readable media of claim 19 , wherein the program instructions are further configured, when executed, to cause a computing system to perform functions comprising:

receiving a user prediction request from a user device;

generating a prediction report based on at least one lookup table of the one or more lookup tables; and

transmitting the prediction report to the user device.

Assignments (2)

Continuity (1)

References Cited (249)