Methods and apparatus for predictive analysis

• US 8,065,250 B2
• Filed: 02/11/2008
• Issued: 11/22/2011
• Est. Priority Date: 02/16/2000
• Status: Expired due to Fees

• Current Assignee: Adaptive Technologies, Inc.
• Original Assignee: Adaptive Technologies, Inc.
• Inventors: Stephens, Christoper R.
• Primary Examiner(s): Gaffin; Jeffrey A
• Assistant Examiner(s): Wong; Lut
• Application Number:
  US12/029,083
• Publication Number:
  US 20080208777A1
• Time in Patent Office: 1,380 Days
• Field of Search: 706/45, 706/12
• US Class Current: 706/45
• CPC Class Codes: G06N 20/00 Machine learning

Abstract

Methods and apparatus for predictive analytics generally comprise one or more artificial agents and an agent factory. An artificial agent may be responsive to at least one of an internal data set and an external data set. Further, an artificial agent may produce a correlation data set relating an outcome data set and at least one of the internal data set and the external data set. In addition, an artificial agent may produce a predictability value corresponding to the correlation data set. The agent factory may be responsive to the outcome data set. Also, the agent factory may produce the artificial agent in response to the outcome data set.

Claims

1. A non-transitory computer-readable medium containing instructions for causing a computer system to process data according to a predictive modeling system, the predictive modeling system comprising:

a plurality of artificial agents responsive to an input data set, wherein;

at least one artificial agent comprises a predictive algorithm;

the input data set comprises at least one intake data set and at least one outcome data set, wherein;

a first intake data set corresponds to a set of factors for a first cohort at an initial state in time; and

a first outcome data set corresponds to a change in the set of factors for the first cohort after a passage of time;

each artificial agent produces a correlation data set relating at least a portion of the first outcome data set with at least a portion of the first intake data set for predicting a future change to a second cohort over a second passage of time based on a second intake data set corresponding to a second set of factors for the second cohort at a second initial state in time; and

each artificial agent produces a predictability value for quantifying an accuracy of the correlation data set produced by the artificial agent; and

an agent factory responsive to the input data set, wherein the agent factory produces the plurality of artificial agents in response to the input data set.

2. A non-transitory computer-readable medium according to claim 1, wherein:

the input data set comprises a second outcome data set corresponding to an actual change in the second set of factors for the second cohort after the second passage of time;

the input data set further comprises a third intake data set corresponding to a set of factors for a third cohort at a third initial state in time;

each artificial agent;

relates at least a portion of the second intake data set with at least a portion of the second outcome data set;

updates the correlation data set based upon the newly related data; and

updates the predictability value in response to the second outcome data set; and

at least one artificial agent updates the estimated outcome data set based on the correlation data set as updated and at least a portion of the third intake data set to predict a future change to the third cohort over a third passage of time.

3. A non-transitory computer readable medium according to claim 1, wherein at least one predictability value is based on a multivariate landscape generated from historical data.

4. A non-transitory computer-readable medium according to claim 1, further comprising an input data module in communication with the agent factory, wherein the input data module at least one of selectively receives, selectively stores, and selectively transmits at least one input data set.

5. A non-transitory computer-readable medium according to claim 1, further comprising a management system in communication with the agent factory, wherein the management system at least one of activates and retires at least one artificial agent according to the at least one artificial agent'"'"'s predictability value.

6. A non-transitory computer-readable medium according to claim 1, further comprising an output data module in communication with the agent factory, wherein the output data module selectively displays at least a portion of at least one estimated outcome data set and at least one predictability value.

7. A computer-implemented method for predictive modeling, comprising:

receiving within an agent factory an input data set, wherein the input data set comprises at least one intake data set and at least one outcome data set, wherein;

a first intake data set corresponds to a set of factors for a first cohort at an initial state in time; and

a first outcome data set corresponds to a change in the set of factors for the first cohort after a passage of time;

generating within the agent factory a plurality of artificial agents, wherein at least one artificial agent comprises a predictive algorithm;

determining within each artificial agent a correlation data set relating at least a portion of the first outcome data set with at least a portion of the first intake data set for predicting a future change to a second cohort over a second passage of time based on a second intake data set corresponding to a second set of factors for the second cohort at a second initial state in time; and

determining within each artificial agent a predictability value for quantifying an accuracy of the correlation data set produced by the artificial agent.

8. A computer-implemented method according to claim 7, further comprising:

updating within at least one artificial agent the estimated outcome data set, wherein;

the input data set comprises a second outcome data set corresponding to an actual change in the second set of factors for the second cohort after the second passage of time;

the input data set further comprises a third intake data set corresponding to a set of factors for a third cohort at a third initial state in time;

each artificial agent;

relates at least a portion of the second intake data set with at least a portion of the second outcome data set;

updates the correlation data set based upon the newly related data; and

updates the predictability value in response to the second outcome data set; and

the estimated outcome data set is updated based on the correlation data set as updated and at least a portion of the third intake data set to predict a future change to the third cohort over a third passage of time.

9. A computer-implemented method according to claim 7, wherein at least one predictability value is based on a multivariate landscape generated from historical data.

10. A computer-implemented method according to claim 7, further comprising:

at least one of selectively receiving, selectively storing, and selectively transmitting at least one input data set via an input data module, wherein the input data module is in communication with the agent factory.

11. A computer-implemented method according to claim 7, further comprising:

at least one of activating and retiring at least one artificial agent according to the predictability value via a management system, wherein the management system is in communication with the agent factory.

12. A computer-implemented method according to claim 7, further comprising:

displaying at least one estimated data set and at least one predictability value via an output data module, wherein the output data module is in communication with the agent factory.

13. A computer-implemented method for candidate screening, comprising:

receiving within an agent factory an input data set, wherein the input data set comprises at least one of an intake data set and an outcome data set, wherein;

a first intake data set corresponds to a set of factors for a first cohort at an initial state in time; and

a first outcome data set corresponds to a change in the set of factors for the first cohort after a passage of time;

generating within the agent factory a plurality of artificial agents, wherein at least one artificial agent comprises a predictive algorithm;

determining within each artificial agent a correlation data set relating at least a portion of the first outcome data set with at least a portion of the first intake data set for predicting a future change to a second cohort over a second passage of time based on a second intake data set corresponding to a second set of factors for the second cohort at a second initial state in time;

determining within each artificial agent a predictability value for quantifying an accuracy of the correlation data set produced by the artificial agent.

14. A computer-implemented method according to claim 13, further comprising:

updating within at least one artificial agent the estimated outcome data set, wherein;

the input data set comprises a second outcome data set corresponding to an actual change in the second set of factors for the second cohort after the second passage of time;

the input data set further comprises a third intake data set corresponding to a set of factors for a third cohort at a third initial state in time;

each artificial agent;

relates at least a portion of the second intake data set with at least a portion of the second outcome data set;

updates the correlation data set based upon the newly related data; and

updates the predictability value in response to the second outcome data set; and

the estimated outcome data set is updated based on the correlation data set as updated and at least a portion of the third intake data set to predict a future change to the third cohort over a third passage of time.

15. A computer-implemented method according to claim 13, wherein at least one predictability value is based on a multivariate landscape generated from historical data.

16. A computer-implemented method according to claim 13, further comprising:

at least one of selectively receiving, selectively storing, and selectively transmitting at least one input data set via an input data module, wherein the input data module is in communication with the agent factory.

17. A computer-implemented method according to claim 13, further comprising:

at least one of activating and retiring at least one artificial agent according to the predictability value via a management system, wherein the management system is in communication with the agent factory.

18. A computer-implemented method according to claim 13, further comprising:

displaying at least one estimated data set and at least one predictability value via an output data module, wherein the output data module is in communication with the agent factory.