Granted Patent B2

US 12,189,933 · App. 17/961,908 · Granted Jan 7, 2025

Inventors: Robert Speare (New York, NY); Dayang Shi (New York, NY); Spencer Lake (New York, NY)

Assignee: Palantir Technologies Inc.

G06F3/04847G06F3/0482G06F9/451G06F16/26G06N20/00

Abstract

Claims (91)

1. A computer system with an improved user interface, the computer system comprising:

one or more processors configured to execute computer readable instructions to cause the computer system to:

generate display data useable for displaying a user interface comprising:

a first input element for receiving one or more optimization constraints;

a second input element for receiving an identifier of a contribution model usable for determining, based at least in part on target parameters associated with each of a plurality of entities, a modeled contribution value for each of the plurality of entities;

a third input element for receiving identifications of grouping parameters; and

a fourth input element for receiving an identifier of a backup model configured to be used for particular entities of the plurality of entities that do not satisfy grouping criteria identified in the contribution model;

optimize, using at least the one or more optimization constraints, the contribution model, the backup model, and the grouping parameters, the target parameters to achieve an optimization target value, wherein the contribution model is utilized for entities of the plurality of entities that satisfy the grouping criteria, and wherein the backup model is used for entities of the plurality of entities that do not satisfy the grouping criteria;

determine, based on the grouping parameters, one or more groups of the plurality of entities;

generate a first graphical visualization comprising a first graph plot that shows, for each different group of the one or more groups of the plurality of entities, a modeled relationship between the modeled contribution value and a change in the target parameter;

cause an update to the user interface to include the first graphical visualization;

receive an input, to the second input element of the user interface, and from a user, indicating a change from the contribution model to a second contribution model;

in response to the input to the second input element of the user interface, cause an update to the user interface to prompt the user to identify updated grouping parameters;

optimize, using at least the one or more optimization constraints, the second contribution model, the backup model, and the updated grouping parameters, the target parameters to achieve an optimization target value, wherein the second contribution model is utilized for entities of the plurality of entities that satisfy the grouping criteria, and wherein the backup model is used for entities of the plurality of entities that do not satisfy the grouping criteria;

determine, based on the updated grouping parameters, one or more updated groups of the plurality of entities;

generate a second graphical visualization comprising a second graph plot that shows, for each different group of the one or more updated groups of the plurality of entities, a modeled relationship between the modeled contribution value and a change in the target parameter; and

cause an update to the user interface to include the second graphical visualization.

2. The computer system of claim 1 , wherein the optimizing comprises:

changing at least some of the target parameters associated the entities, wherein the target parameters are changed to be within the one or more optimization constraints;

determining an entity contribution value for each of the entities using the contribution model; and

determining contributions of the target parameters associated with each of the entities toward the optimization target value based at least in part on the entity contribution values for each of the entities.

3. The computer system of claim 1 , wherein the one or more processors are configured to execute computer readable instructions to further cause the computer system to:

cache values for the target parameters associated with each of the entities; and

optimize the cached values for the target parameters within the one or more optimization constraints to achieve the optimization target value.

4. The computer system of claim 1 , wherein the one or more processors are configured to execute computer readable instructions to further cause the computer system to:

translate the optimization constraints into an optimization problem; and

cache data associated with the optimization problem to speed up iterations of an optimization algorithm.

5. The computer system of claim 1 , wherein the one or more processors are configured to execute computer readable instructions to further cause the computer system to:

generate display data useable for displaying a modeling user interface comprising:

fifth input element for receiving indication of a historical data file;

an option to generate a model file based on the historical data file; and

a sixth input element to specify where to save the model file.

6. The computer system of claim 1 , wherein the user interface further comprises:

a fifth input element for specifying a type for the contribution model; and

a display area configured to show different group combinations based on the grouping parameters and coefficients determined for the contribution model, wherein the grouping parameters include at least two grouping parameters selected by a user, and wherein the coefficients are determined based at least on the specified type for the contribution model.

7. The computer system of claim 6 , wherein:

the display area is configured to show different groups of entities, and

a combination of values of the grouping parameters is shown for each group in the display area.

8. The computer system of claim 1 , wherein the user interface further comprises:

a fifth input element for receiving a selection of an optimization step; and

a second graphical visualization showing a result achieved during the optimization step.

9. The computer system of claim 1 , wherein the user interface further comprises:

a second graphical visualization showing a distribution of optimized values for the target parameter associated with each of the entities.

10. The computer system of claim 9 , wherein the visualization is a frontier plot.

11. The computer system of claim 1 , wherein the user interface further comprises:

a second display area showing statistics about optimized values for the target parameter associated with each of the entities.

12. The computer system of claim 1 , wherein the user interface further comprises:

a fifth input element for a file path for saving optimized values for the target parameter.

13. The computer system of claim 1 , wherein the user interface further comprises:

a fifth input element for receiving an identification of a data file comprising:

identifiers of the plurality of entities; and

the target parameters associated with each of the entities; and

a sixth input element for receiving the optimization target value.

14. A computer-implemented method comprising:

by one or more processors configured to execute computer readable instructions:

generating display data useable for displaying a user interface comprising:

a first input element for receiving one or more optimization constraints;

a second input element for receiving an identifier of a contribution model useable for determining, based at least in part on target parameters associated with each of a plurality of entities, a modeled contribution value for each of the plurality of entities;

a third input element for receiving identifications of grouping parameters; and

a fourth input element for receiving an identifier of a backup model configured to be used for particular entities of the plurality of entities that do not satisfy grouping criteria identified in the contribution model;

optimizing, using at least the one or more optimization constraints, the contribution model, the backup model, and the grouping parameters, the target parameters to achieve an optimization target value, wherein the contribution model is utilized for entities of the plurality of entities that satisfy the grouping criteria, and wherein the backup model is used for entities of the plurality of entities that do not satisfy the grouping criteria;

determining, based on the grouping parameters, one or more groups of the plurality of entities;

generating a first graphical visualization comprising a first graph plot that shows, for each different group of the one or more groups of the plurality of entities, a modeled relationship between the modeled contribution value and a change in the target parameter;

causing an update to the user interface to include the first graphical visualization;

receiving an input, to the second input element of the user interface, and from a user, indicating a change from the contribution model to a second contribution model;

in response to the input to the second input element of the user interface, causing an update to the user interface to prompt the user to identify updated grouping parameters;

optimizing, using at least the one or more optimization constraints, the second contribution model, the backup model, and the updated grouping parameters, the target parameters to achieve an optimization target value, wherein the second contribution model is utilized for entities of the plurality of entities that satisfy the grouping criteria, and wherein the backup model is used for entities of the plurality of entities that do not satisfy the grouping criteria;

determining, based on the updated grouping parameters, one or more updated groups of the plurality of entities;

generating a second graphical visualization comprising a second graph plot that shows, for each different group of the one or more updated groups of the plurality of entities, a modeled relationship between the modeled contribution value and a change in the target parameter; and

causing an update to the user interface to include the second graphical visualization.

15. The computer-implemented method of claim 14 , wherein the optimizing comprises:

changing at least some of the target parameters associated the entities, wherein the target parameters are changed to be within the one or more optimization constraints;

determining an entity contribution value for each of the entities using the contribution model; and

determining contributions of the target parameters associated with each of the entities toward the optimization target value based at least in part on the entity contribution values for each of the entities.

16. The computer-implemented method of claim 14 further comprising:

by the one or more processors configured to execute computer readable instructions:

caching values for the target parameters associated with each of the entities; and

optimizing the cached values for the target parameters within the one or more optimization constraints to achieve the optimization target value.

17. The computer-implemented method of claim 14 further comprising:

by the one or more processors configured to execute computer readable instructions:

translating the optimization constraints into an optimization problem; and

caching data associated with the optimization problem to speed up iterations of an optimization algorithm.

18. The computer-implemented method of claim 14 further comprising:

by the one or more processors configured to execute computer readable instructions:

generating display data useable for displaying a modeling user interface comprising:

a fifth input element for receiving indication of a historical data file;

an option to generate a model file based on the historical data file; and

a sixth input element to specify where to save the model file.

19. The computer-implemented method of claim 14 , wherein the user interface further comprises:

a fifth input element for specifying a type for the contribution model; and

a display area configured to show different group combinations based on the grouping parameters and coefficients determined for the contribution model, wherein the grouping parameters include at least two grouping parameters selected by a user, and wherein the coefficients are determined based at least on the specified type for the contribution model.

Assignments (1)

Continuity (4)

References Cited (353)