SYSTEMS AND METHODS FOR FORECASTING USING CUSTOMER PREFERENCE PROFILES

• US 20140324534A1
• Filed: 04/30/2013
• Published: 10/30/2014
• Est. Priority Date: 04/30/2013
• Status: Abandoned Application

• Current Assignee: Caterpillar Incorporated
• Original Assignee: Caterpillar Incorporated
• Inventors: GRICHNIK, Anthony James, KERSH, Thad Breton, SESKIN, Michael
• Application Number:
  US13/873,698
• Publication Number:
  US 20140324534A1
• Time in Patent Office: Days
• US Class Current: 705/7.31
• CPC Class Codes: G06Q 30/0202 Market predictions or forecasting for commercial activities

Abstract

A computer-implemented method for forecasting characteristics of a target item is disclosed. The method may include determining a forecast function representing characteristics of the target item. The forecast function may include one or more generic functions and one or more customer-specific functions. Each one of the generic functions may be uniform for a plurality of customers within one or more customer groups and may have one or more generic function variables. Each one of the customer-specific functions may be specifically configured for a corresponding customer group and may have one or more customer-specific function variables. The method may also include determining a data value for each one of the generic function variables and the customer-specific function variables by using a genetic algorithm. The method may further include forecasting the characteristics of the target item by using the forecast function and the determined data values.

Claims

1. A computer-implemented method for forecasting characteristics of a target item comprising:

determining a forecast function representing characteristics of the target item, wherein the forecast function includes one or more generic functions and one or more customer-specific functions, each one of the generic functions is uniform for a plurality of customers within one or more customer groups and has one or more generic function variables, and each one of the customer-specific functions is specifically configured for a corresponding customer group and has one or more customer-specific function variables;

determining, by one or more processors, a data value for each one of the generic function variables and the customer-specific function variables by using a genetic algorithm; and

forecasting, by the one or more processors, the characteristics of the target item by using the forecast function and the determined data values.

2. The computer-implemented method of claim 1, further including:

receiving a first set of historical data regarding orders placed by each one of the plurality of customers for the target item as a replacement or service part for one or more machines;

receiving a second set of historical data regarding machine conditions of the one or more machines;

determining a customer preference profile for each one of the customers based on the first and second sets of historical data, wherein the customer preference profile represents a relationship between the orders placed by the customer and the machine conditions of the machine under which the customer places the orders;

identifying one or more customer groups each including one or more of the customers having similar customer preference profiles; and

determining the customer-specific function for each one of the customer groups.

3. The computer-implemented method of claim 2, wherein the customer-specific function represents a forecast of the characteristics of the target item for the corresponding customer group.

4. The computer-implemented method of claim 2, wherein the machine condition includes one or more of machine age, maintenance and repair history of the machine, and current state of the machine.

5. The computer-implemented method of claim 4, wherein the current state of the machine includes one or more of temperature, pressure, location, speed, and fuel consumption rate of the machine.

6. The computer-implemented method of claim 1, wherein each one of the customer-specific functions has a relative weighting variable, and the forecast function is represented as:

7. The computer-implemented method of claim 1, wherein determining the data value for each one of the generic function variables and the customer-specific function variables by using the genetic algorithm further includes:

generating one or more chromosomes having a data value for each of the generic function variables and the customer-specific function variables;

determining a chromosome value for at least one of the chromosomes based on a goal function including one or more measureable business goals; and

selecting a chromosome from among the one or more chromosomes based on the chromosome value.

8. The computer-implemented method of claim 7, wherein the goal function includes at least one measureable business goal selected from the group consisting of profit, return on net assets, inventory turns, and service level.

9. The computer-implemented method of claim 1, wherein each one of the customer-specific functions has a relative weighting variable, and the method further including:

determining a data value for each one of the relative weighting variables by using the genetic algorithm.

10. A characteristic forecasting system comprising:

a processor; and

a memory module configured to store instructions, that, when executed, enable the processor to;

determine a forecast function representing characteristics of the target item, wherein the forecast function includes one or more generic functions and one or more customer-specific functions, each one of the generic functions is uniform for a plurality of customers within one or more customer groups and has one or more generic function variables, and each one of the customer-specific functions is specifically configured for a corresponding customer group and has one or more customer-specific function variables;

determine a data value for each one of the generic function variables and the customer-specific function variables by using a genetic algorithm; and

forecast the characteristics of the target item by using the forecast function and the determined data values.

11. The system of claim 10, the instructions stored in the memory module further enabling the processor to:

receive a first set of historical data regarding orders placed by each one of the plurality of customers for the target item as a replacement or service part for one or more machines;

receive a second set of historical data regarding machine conditions of the one or more machines;

determine a customer preference profile for each one of the customers based on the first and second sets of historical data, wherein the customer preference profile represents a relationship between the orders placed by the customer and the machine conditions of the machine under which the customer places the orders;

identify one or more customer groups each including one or more of the customers having similar customer preference profiles; and

determine the customer-specific function for each one of the customer groups.

12. The system of claim 11, wherein the customer-specific function represents a forecast of the characteristics of the target item for the corresponding customer group.

13. The system of claim 11, wherein the machine condition includes one or more of machine age, maintenance and repair history of the machine, and current state of the machine.

14. The system of claim 13, wherein the current state of the machine includes one or more of temperature, pressure, location, speed, and fuel consumption rate of the machine.

15. The system of claim 10, wherein each one of the customer-specific functions has a relative weighting variable, and the forecast function is represented as:

16. The system of claim 10, the instructions stored in the memory module further enabling the processor to:

generate one or more chromosomes having a data value for each of the generic function variables and the customer-specific function variables;

determine a chromosome value for at least one of the chromosomes based on a goal function including one or more measureable business goals; and

select a chromosome from among the one or more chromosomes based on the chromosome value.

17. The system of claim 16, wherein the goal function includes at least one measureable business goal selected from the group consisting of profit, return on net assets, inventory turns, and service level.

18. The system of claim 10, wherein each one of the customer-specific functions has a relative weighting variable, and the instructions stored in the memory module further enabling the processor to:

determine a data value for each one of the relative weighting variables by using the genetic algorithm.

19. A non-transitory computer-readable storage device storing instructions for forecasting characteristics of a target item, the instructions causing one or more computer processors to perform operations comprising:

determining a forecast function representing characteristics of the target item, wherein the forecast function includes one or more generic functions and one or more customer-specific functions, each one of the generic functions is uniform for a plurality of customers within one or more customer groups and has one or more generic function variables, and each one of the customer-specific functions is specifically configured for a corresponding customer group and has one or more customer-specific function variables;

determining a data value for each one of the generic function variables and the customer-specific function variables by using a genetic algorithm; and

forecasting the characteristics of the target item by using the forecast function and the determined data values.

20. The computer-readable storage device of claim 19, the instructions further causing the one or more computer processors to perform operations including:

receiving a first set of historical data regarding orders placed by each one of the plurality of customers for the target item as a replacement or service part for one or more machines;

receiving a second set of historical data regarding machine conditions of the one or more machines;

determining a customer preference profile for each one of the customers based on the first and second sets of historical data, wherein the customer preference profile represents a relationship between the orders placed by the customer and the machine conditions of the machine under which the customer places the orders;

identifying one or more customer groups each including one or more of the customers having similar customer preference profiles; and

determining the customer-specific function for each one of the customer groups.