Method for stabilized tuning of demand models

• US 6,078,893 A
• Filed: 05/21/1998
• Issued: 06/20/2000
• Est. Priority Date: 05/21/1998
• Status: Expired due to Term

• Current Assignee: SAP AG (SAP SE)
• Original Assignee: Khimetrics, Inc. (SAP SE)
• Inventors: Ouimet, Kenneth J., Chaubal, Charu V.
• Primary Examiner(s): Swann, Tod R.
• Assistant Examiner(s): Irshadullah, M.
• Application Number:
  US09/083,647
• Time in Patent Office: 761 Days
• Field of Search: 705/1, 705/7, 705/10, 705/14, 235/381
• US Class Current: 705/7.31
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0202 Market predictions or forecasting for commercial activities

Abstract

A method for tuning a demand model in manner that is stable with respect to fluctuations in the sales history used for the tuning is provided. A market model is selected, which predicts how a subset of the parameters in the demand model depends upon information external to the sales history; this model may itself have a number of parameters. An effective figure-of-merit function is defined, consisting of a standard figure-of-merit function based upon the demand model and the sales history, plus a function that attains a minimum value when the parameters of the demand model are closest to the predictions of the market model. This effective figure-of-merit function is minimized with respect to the demand model and market model parameters. The resulting demand model parameters conform to the portions of the sales history data that show a strong trend, and conform to the external market information when the corresponding portions of the sales history data show noise.

Claims

1. A computer-implemented method for tuning a demand model to a sales history that is stable with respect to random fluctuations in said sales history, said method comprising the steps of:

providing a data storage having sales history data;

selecting a demand model, said demand model depending on a plurality of parameters including at least one demand parameter;

selecting a market model to describe how a subset of said at least one demand parameter is expected to vary according to external market information, said market model depending on at least one market parameter, which is external to said sales history data;

constructing a an effective figure-of merit function from said demand model and said market model, said effective figure-of-merit function having a plurality of parameters including said at least one demand parameter and said at least one market parameter; and

minimizing said effective figure-of-merit function with respect to said at least one demand parameter and said at lest one market parameter, thereby attaining an optimal value for said at least one demand parameter that is stable with respect to random fluctuations in said sales history data.

2. The computer-implemented method of claim 1 wherein said demand model predicts the demand for at least one item.

3. The computer-implemented method of claim 1 wherein said demand model predicts the demand for at least one service.

4. The computer-implemented method of claim 1 wherein said demand model is a function having the form:

space="preserve" listing-type="equation">q=q({X};

{D})

where {D} is said at least one demand parameter and {X} is a set of all variable for said at least one item or service.

5. The computer-implemented method of claim 1 wherein said market model is a function having the form:

space="preserve" listing-type="equation">M=M({g},{S})

where {g} is said subset of said at least one demand parameter and {S} is said at least one market parameter.

6. The computer-implemented method of claim 1 wherein said market model is defined such that its value is minimized as said subset of said at least one demand parameter is optimized with respect to said external market information.

7. The computer-implemented method of claim 6 further comprising the step of:

constructing a figure-of-merit function from said demand model such that said figure-of-merit function attains a minimum value when said at least one demand parameter is optimized with respect to said sales history data.

8. The computer-implemented method of claim 7 wherein said step of constructing said effective figure-of-merit function further includes the step of:

combining said figure-of-merit function with said market model to form said effective figure-of-merit function.

9. The computer-implemented method of claim 1 further comprising the step of storing said optimal value for said at least one demand parameter in an external data storage, whereby said optimal value can be used in another routine.

10. A computer-readable storage medium, for use in a computer having a storage device for storing sales history data, said computer-readable storage medium for storing program code for causing the computer to perform the steps of:

selecting a demand model to predict the demand of a plurality of parameters including at least one demand parameter;

selecting a market model to describe how a subset of said at least one demand parameter is expected to vary according to external market information, said market model depending on at least one market parameter, which is external to said sales history data;

combining said demand model and said market model to from an effective figure-of-merit function; and

minimizing said effective figure-of-merit function with respect to said at least one demand parameter and said at least one market parameter, thereby attaining an optimal value for said at least one demand parameter that is stable with respect to random fluctuations in said sales history data.

11. The computer-readable storage medium of claim 10 wherein said demand model is a function having the form:

space="preserve" listing-type="equation">q=q({X};

{D})

where {D} is said at least one demand parameter and {X} is a set of all variable for said plurality of items.

12. The computer-readable storage medium of claim 10 wherein said market model is a function having the form:

space="preserve" listing-type="equation">M=M({g},{S})

where {g} is said subset of said at least one demand parameter and {S} is said at least one market parameter.

13. The computer-readable storage medium of claim 10 wherein said market model is defined such that its value approaches zero as said subset of said at least one demand parameter is optimized with respect to said external market information.

14. The computer-readable storage medium of claim 10 further comprising program code for causing the computer to perform the step of:

constructing a figure-of-merit function from said demand model such that said figure-of-merit function attains a minimum value when said at least one demand parameter is optimized with respect to said sales history data;

wherein said effective figure-of-merit function is formed by combining said figure-of-merit function with said market model.

15. The computer-readable storage medium of claim 10 further comprising program code for causing the computer to perform the step of:

storing said optimized value for said at least one demand parameter in a storage device, whereby said optimized value can be used in another routine.

16. A system for tuning a demand model to a sales history that stabilizes said demand model with respect to fluctuations in said sales history, said system for use with a storage device for storing sales history data, said system comprising:

means for selecting a demand model and a market model, said demand model depending on a plurality of parameters including at least one demand parameter;

wherein said market model describes how a subset of said at least one demand parameter is expected to vary according to external market information, said market model depending on at least one market parameter that is external to said sales history;

means for combining said demand model and said market model to form an effective figure-of-merit function; and

means for optimizing said at least one demand parameter in said effective figure-of-merit function that is stable with respect to random fluctuations in said sales history data.

17. The system of claim 16 further comprising means for accessing said external market data.

18. The system of claim 16 further comprising:

means for storing said optimized value of said at least one demand parameter, whereby said optimized value can be used in another routine.