Infrastructure and Architecture for Development and Execution of Predictive Models

• US 20090259664A1
• Filed: 04/09/2008
• Published: 10/15/2009
• Est. Priority Date: 04/09/2008
• Status: Active Grant

• Current Assignee: American Express Travel Related Services Company, Inc. (American Express Company)
• Original Assignee: American Express Travel Related Services Company, Inc. (American Express Company)
• Inventors: Oralevich, Mickey S., SACHETI, Sandeep, Murthy, Narasimha, Thomas, Deep
• Granted Patent: US 8,341,166 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/13 File access structures, e.g. distributed indices arrangements of input from, or output to, record carriers G06F3/06
• CPC Class Codes: G06F 16/148 File search processing
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06N 5/04 Inference or reasoning models
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0201 Market modelling; Market analysis; Collecting market data

Abstract

A system that enables development and execution of predictive models comprises a centralized data management system, a data extraction tool a model validation tool and a model execution tool. In embodiments, a data management system includes a data management server that can be accessed via a web browser that stores data in the form of a flat file. An extraction tool extracts data. A model validation tool validates a model by scoring an analytical environment data set and a production environment data set. A model execution tool allows a user to select when and how often a model is scored.

Claims

1. A system for enabling development and execution of predictive models comprising:

a centralized data management system accessible via a web browser operable to receive data from a plurality of data feeds, wherein the centralized data management subsystem is further operable to store the data to flat files;

a data extraction tool operable to extract data from the data management system, based at least in part on the contents of an uploaded file thereby allowing a user to create a model;

a model validation tool operable to validate the model;

a model execution the operable to score customers on the model, wherein the model execution tool allows a user to select the frequency at which the model is scored.

2. The system of claim 1, wherein a flat file comprises a SAS data set.

3. The system of claim 1, wherein the model validation tool validates a model at least in part by:

scoring an analytical environment data set;

scoring a production environment data set; and

comparing scores generated by the analytical environment data set to scores generated by the production environment data set.

4. The system of claim 3, wherein the results of the comparison between scores generated by the analytical environment data set and scores generated by the production environment are produced in a report.

5. The system of claim 1, wherein the model validation tool validates the model at least in part by using a pre-defined test case.

6. The system of claim 1, wherein the model execution tool includes an expiry date function that allows a user to specify how long the model will run.

7. The system of claim 1, wherein the model execution tool is further operable to organize variables into files based on the number of models that use a particular variable.

8. A method for enabling development and execution of predictive models comprising:

receiving data from a plurality of data feeds, the data feeds comprising a set of variables;

storing the data to flat files;

extracting data from the flat files, based at least in part on the contents of an uploaded file thereby allowing a user to create a model;

validating the model;

specifying the frequency at which the validated model is scored; and

scoring customers data on the validated model.

9. The method of claim 8, wherein a flat file comprises a SAS data set.

10. The method of claim 8, wherein said validating comprises:

scoring an analytical environment data set;

scoring a production environment data set; and

comparing scores generated by the analytical environment data set to scores generated by the production environment data set.

11. The method of claim 10, wherein the results of the comparison between scores generated by the analytical environment data set and scores generated by the production environment are produced in a report.

12. The method of claim 8, wherein said validating further comprises validating the model based at least in part on a pre-defined test case.

13. The method of claim 8, wherein said specifying the frequency at which the validated model is scored comprises specifying how long the model will be scored.

14. The method of claim 8, further comprising:

organizing variables into files based on the number of models that use a particular variable.

15. A computer program product comprising a computer usable medium including control logic stored therein, the control logic enabling the development and execution of predictive models by performing a method, the method comprising:

receiving data from a plurality of data feeds, the data feeds comprising a set of variables;

storing the data to flat files;

extracting data from the flat files, based at least in part on the contents of an uploaded file thereby allowing a user to create a model;

validating the model;

specifying the frequency at which the validated model is scored; and

scoring customers data on the validated model.

16. The computer program product of claim 15, wherein a flat file comprises a SAS data set.

17. The computer program product of claim 15, wherein said validating comprises:

scoring an analytical environment data set;

scoring a production environment data set; and

comparing scores generated by the analytical environment data set to scores generated by the production environment data set.

18. The computer program product of claim 17, wherein the results of the comparison between scores generated by the analytical environment data set and scores generated by the production environment are produced in a report.

19. The computer program product of claim 15, wherein said validating further comprises validating the model based at least in part on a pre-defined test case.

20. The computer program product of claim 15, wherein the method further comprises:

organizing variables into files based on the number of models that use a particular variable.