Financial product application pull-through system

• US 8,515,841 B2
• Filed: 09/10/2010
• Issued: 08/20/2013
• Est. Priority Date: 09/10/2010
• Status: Active Grant

• Current Assignee: Bank of America Corp.
• Original Assignee: Bank of America Corp.
• Inventors: Thalken, Jason, Lamarque, Jeffrey
• Primary Examiner(s): Shaikh, Mohammad Z
• Application Number:
  US12/879,857
• Publication Number:
  US 20120066104A1
• Time in Patent Office: 1,075 Days
• Field of Search: 705 35- 45
• US Class Current: 705/35
• CPC Class Codes: G06Q 40/00 Finance; Insurance; Tax strategies; Processing of corporate or income taxes
• CPC Class Codes: G06Q 40/06 Asset management; Financial planning or analysis

Abstract

A method evaluates a prediction set of financial product applications. The method builds a multi-dimensional observation grid of observation grid points, each corresponding to at least one past financial product application in an observation set and each populated with the one or more variable values and the observation result value associated with the past financial product application. The method includes building a multi-dimensional prediction grid comprising a plurality of prediction grid points, each corresponding to a present financial product application in the prediction set and populated with one or more variable values of the present financial product application, each also corresponding to an observation grid point and assigning a prediction result value for each prediction grid point. In some embodiments, the method defines an observation set and a prediction set comprising a past and a present financial product application, respectively, each application including two or more corresponding variable values.

Claims

1. A method for evaluating a prediction set of financial product applications, the method comprising:

defining, by a processing device, an observation set comprising at least one past financial product application, wherein the past financial product application comprises one or more corresponding variable values and an observation result value;

building, by the processing device, a multi-dimensional observation grid comprising a plurality of observation grid points, each observation grid point corresponding to at least one past financial product application in the observation set and each observation grid point populated with the one or more variable values and the observation result value associated with the at least one past financial product application;

defining, by the processing device, a prediction set comprising at least one present financial product application, wherein each financial product application comprises one or more corresponding variable values;

building, by the processing device, a multi-dimensional prediction grid comprising a plurality of prediction grid points, each prediction grid point corresponding to at least one present financial product application in the prediction set and each prediction grid point populated with one or more variable values of the at least one present financial product application, wherein each of the prediction grid points also corresponds to an observation grid point of the observation grid;

assigning, by the processing device, a prediction result value for each prediction grid point based at least in part on an observation result value populated in the observation grid point corresponding with the prediction grid point;

compressing, by the processing device, the multi-dimensional prediction grid into a single-dimensional grid wherein each prediction grid point is populated consecutively into the single-dimensional grid from lowest prediction result value to highest prediction result value;

stepping through, by the processing device, the single-dimensional grid from lowest prediction result value to highest prediction result value to identify a plurality of grid sections; and

assigning, by the processing device, a prediction score to each grid section, wherein the prediction score assigned to each successive grid section is incrementally assigned from a set of prediction scores.

2. The method of claim 1, wherein the multi-dimensional prediction grid is a three dimensional grid wherein each prediction grid point is populated with two variable values and one prediction result value.

3. The method of claim 1, wherein the plurality of grid sections are identified based at least in part on a predetermined percentage of the number of prediction grid points having been stepped through.

4. The method of claim 3, wherein the predetermined percentage of the number of prediction grid points stepped through is five percent and the plurality of grid sections comprises twenty grid sections.

5. The method of claim 4, wherein the prediction score assigned to each successive grid section is incrementally assigned from a set of prediction scores comprising scores of zero to nineteen, and wherein the prediction score assigned to each successive grid section is incrementally assigned starting with zero and ending with nineteen.

6. The method of claim 1, further comprising:

calculating, by the processing device, the observation result value populated with each observation grid point, the calculating comprising;

dividing, by the processing device, the number of past financial product applications in the observation set having a negative result status corresponding with the observation grid point by the total number of past financial product applications in the observation set corresponding with the observation grid point, thereby resulting in the observation result value for the observation grid point.

7. The method of claim 6, wherein the negative result status comprises the past financial product application having been canceled by a customer.

8. The method of claim 6, wherein the negative result status comprises the past financial product application having been denied by a financial institution.

9. The method of claim 1, further comprising:

dividing, by the processing device, the prediction set into groups based at least in part on one or more variables associated with the one or more variable values;

building, by the processing device, a multi-dimensional prediction grid for each group, each multi-dimension prediction grid comprising a plurality of prediction grid points; and

assigning, by the processing device, a prediction result value for each prediction grid point for each multi-dimensional prediction grid for each group.

10. The method of claim 1, further comprising:

initiating production, by the processing device, of a list of the present financial product applications of the prediction set, the list including the prediction result value assigned to the prediction grid point corresponding to each present financial product application.

11. A system for evaluating a prediction set of financial product applications, the system comprising:

a processing device configured for;

defining an observation set comprising at least one past financial product application, wherein the past financial product application comprises two or more corresponding variable values and an observation result value;

building a multi-dimensional observation grid comprising a plurality of observation grid points, each observation grid point corresponding to at least one past financial product application in the observation set and each observation grid point populated with the one or more variable values and the observation result value associated with the at least one past financial product application;

defining a prediction set comprising at least one present financial product application, wherein each financial product application comprises one or more corresponding variable values;

building a multi-dimensional prediction grid comprising a plurality of prediction grid points, each prediction grid point corresponding to at least one present financial product application in the prediction set and each prediction grid point populated with one or more variable values of the at least one present financial product application, wherein each of the prediction grid points also corresponds to an observation grid point of the observation grid;

assigning a prediction result value for each prediction grid point based at least in part on an observation result value populated in the observation grid point corresponding with the prediction grid point;

compressing the multi-dimensional grid into a single-dimensional grid wherein each prediction grid point is populated consecutively into the single-dimensional grid from lowest prediction result value to highest prediction result value;

stepping through the single-dimensional grid from lowest prediction result value to highest prediction result value to identify a plurality of grid sections; and

assigning a prediction score to each grid section, wherein the prediction score assigned to each successive grid section is incrementally assigned from a set of prediction scores.

12. The system of claim 11, wherein the multi-dimensional prediction grid is a three dimensional grid wherein each prediction grid point is populated with two variable values and one prediction result value.

13. The system of claim 11, wherein the plurality of grid sections are identified based at least in part on a predetermined percentage of the number of prediction grid points having been stepped through.

14. The system of claim 13, wherein the predetermined percentage of the number of prediction grid points stepped through is five percent and the plurality of grid sections comprises twenty grid sections.

15. The system of claim 14, wherein the prediction score assigned to each successive grid section is incrementally assigned from a set of prediction scores comprising scores of zero to nineteen, and wherein the prediction score assigned to each successive grid section is incrementally assigned starting with zero and ending with nineteen.

16. The system of claim 11, wherein the processing device is further configured for:

calculating the observation result value populated with each observation grid point, the calculating comprising;

dividing, by the processing device, the number of past financial product applications in the observation set having a negative result status corresponding with the observation grid point by the total number of past financial product applications in the observation set corresponding with the observation grid point, thereby resulting in the observation result value for the observation grid point.

17. The system of claim 16, wherein the negative result status comprises the past financial product application having been canceled by a customer.

18. The system of claim 16, wherein the negative result status comprises the past financial product application having been denied by a financial institution.

19. The system of claim 11, wherein the processing device is further configured for:

dividing the prediction set into groups based at least in part on one or more variables associated with the one or more variable values;

building a multi-dimensional prediction grid for each group, each multi-dimension prediction grid comprising a plurality of prediction grid points; and

assigning a prediction result value for each prediction grid point for each multi-dimensional prediction grid for each group.

20. The system of claim 11, wherein the processing device is further configured for:

initiating production of a list of the present financial product applications of the prediction set, the list including the prediction result value assigned to the prediction grid point corresponding to each present financial product application.

21. A computer program product comprising a non-transient computer-readable medium comprising instructions for evaluating a prediction set of financial product applications, the instructions comprising:

instructions for defining an observation set comprising at least one past financial product application, wherein the past financial product application comprises two or more corresponding variable values and an observation result value;

instructions for building a multi-dimensional observation grid comprising a plurality of observation grid points, each observation t financial product application in the observation set and each observation grid point populated with the one or more variable values and the observation result value associated with the at least one past financial product application;

instructions for defining a prediction set comprising at least one present financial product application, wherein each financial product application comprises one or more corresponding variable values;

instructions for building a multi-dimensional prediction grid comprising a plurality of prediction grid points, each prediction grid point corresponding to at least one present financial product application in the prediction set and each prediction grid point populated with one or more variable values of the at least one present financial product application, wherein each of the prediction grid points also corresponds to an observation grid point of the observation;

instructions for assigning a prediction result value for each prediction grid point based at least in part on an observation result value populated in the observation grid point corresponding with the prediction grid point;

instructions for compressing the multi-dimensional grid into a single-dimensional grid wherein each prediction grid point is populated consecutively into the single-dimensional grid from lowest prediction result value to highest prediction result value;

instructions for stepping through the single-dimensional grid from lowest prediction result value to highest prediction result value to identify a plurality of grid sections; and

instructions for assigning a prediction score to each grid section, wherein the prediction score assigned to each successive grid section is incrementally assigned from a set of prediction scores.

22. The computer program product of claim 21, wherein the multi-dimensional prediction grid is a three dimensional grid wherein each prediction grid point is populated with two variable values and one prediction result value.

23. The computer program product of claim 21, wherein the plurality of grid sections are identified based at least in part on a predetermined percentage of the number of prediction grid points having been stepped through.

24. The computer program product of claim 23, wherein the predetermined percentage of the number of prediction grid points stepped through is five percent and the plurality of grid sections comprises twenty grid sections.

25. The computer program product of claim 24, wherein the prediction score assigned to each successive grid section is incrementally assigned from a set of prediction scores comprising scores of zero to nineteen, and wherein the prediction score assigned to each successive grid section is incrementally assigned starting with zero and ending with nineteen.

26. The computer program product of claim 21, wherein the instructions further comprise:

instructions for calculating the observation result value populated with each observation grid point, the instructions for calculating comprising;

instructions for dividing the number of past financial product applications in the observation set having a negative result status corresponding with the observation grid point by the total number of past financial product applications in the observation set corresponding with the observation grid point, thereby resulting in the observation result value for the observation grid point.

27. The computer program product of claim 26, wherein the negative result status comprises the past financial product application having been canceled by a customer.

28. The computer program product of claim 26, wherein the negative result status comprises the past financial product application having been denied by a financial institution.

29. The computer program product of claim 21, wherein the instructions further comprise:

instructions for dividing the prediction set into groups based at least in part on one or more variables associated with the one or more variable values;

instructions for building a multi-dimensional prediction grid for each group, each multi-dimension prediction grid comprising a plurality of prediction grid points; and

instructions for assigning a prediction result value for each prediction grid point for each multi-dimensional prediction grid for each group.

30. The computer program product of claim 21, wherein the instructions further comprise:

instructions for initiating production of a list of the present financial product applications of the prediction set, the list including the prediction result value assigned to the prediction grid point corresponding to each present financial product application.

31. A method for evaluating a prediction set of financial product applications, the method comprising:

defining, by a processing device, an observation set comprising at least one past financial product application, wherein the past financial product application comprises one or more corresponding variable values and an observation result value;

defining, by the processing device, a prediction set comprising at least one present financial product application, wherein each financial product application comprises one or more corresponding variable values;

building, by the processing device, a multi-dimensional observation grid comprising a plurality of observation grid points, each observation grid point corresponding to at least one past financial product application in the observation set and each observation grid point populated with the one or more variable values and the observation result value associated with the at least one past financial product application;

building, by the processing device, a multi-dimensional prediction grid comprising a plurality of prediction grid points, each prediction grid point corresponding to at least one present financial product application in the prediction set and each prediction grid point populated with one or more variable values of the at least one present financial product application, each of the prediction grid points also corresponding to an observation grid point of the observation grid;

assigning, by the processing device, a prediction result value for each prediction grid point based at least in part on the observation result value populated in the observation grid point corresponding with the prediction grid point;

compressing the multi-dimensional prediction grid into a single-dimensional grid wherein each prediction grid point is populated consecutively into the single-dimensional grid from lowest prediction result value to highest prediction result value;

stepping through the single-dimensional grid from lowest prediction result value to highest prediction result value to identify a plurality of grid sections; and

assigning a prediction score to each grid section, wherein the prediction score assigned to each successive grid section is incrementally assigned from a set of prediction scores.