Method for estimating a used vehicle's market value

• US 8,005,684 B1
• Filed: 06/29/2000
• Issued: 08/23/2011
• Est. Priority Date: 06/29/2000
• Status: Active Grant

• Current Assignee: Ford Global Technologies, LLC (Ford Motor Company)
• Original Assignee: Ford Global Technologies, LLC (Ford Motor Company)
• Inventors: Lu, Yi, Cheng, Jie, Peng, Rose
• Primary Examiner(s): Mooneyham; Janice A.
• Assistant Examiner(s): Fisher; Michael J
• Application Number:
  US09/607,069
• Time in Patent Office: 4,072 Days
• Field of Search: 705/1.1, 705/306, 708/134
• US Class Current: 705/1.1
• CPC Class Codes: G06Q 10/06 Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
• CPC Class Codes: G06Q 30/0278 Product appraisal
• CPC Class Codes: G06Q 30/06 Buying, selling or leasing transactions

Abstract

A method for estimating a used vehicle'"'"'s market value is disclosed. The method searches for a set of comparable vehicles in an existing historical database of used vehicles. Estimations of the used vehicles are established based on each comparable vehicle (neighbor) in the set of comparable vehicles by adjusting their resale value in accordance with the individual differences found between the neighbor(s) and the used vehicle whose market value is to be estimated/predicted (the target vehicle). Finally, a distance-weighted average of all such estimations is calculated to arrive at a market value estimation for the target vehicle.

Claims

1. A computer-implemented method for estimating market value of a used vehicle, the method comprising:

A) electronically receiving data from a nearest neighbor database consisting of a number K of used vehicle nearest neighbor records, each used vehicle nearest neighbor record comprising resale information and a plurality of used vehicle features, at least one target used vehicle record comprised of a plurality of used vehicle features, at least one constraint for determining a neighbor relationship between a pair of used vehicles, and a neighborhood distance function for determining a distance between a pair of used vehicles, the number K is iteratively selected for estimation accuracy based on a historical database of N used vehicle records; and

B) electronically determining an estimated value for the at least one target used vehicle based on the data from the nearest neighbor database, the at least one target used vehicle record, the at least one constraint, and the neighborhood distance function,

wherein the estimated value of the at least one target used vehicle is relied upon by individuals to at least price used vehicles for resale.

2. The method of claim 1, wherein determining step B) is comprised of:

B1) for each used vehicle nearest neighbor record in the nearest neighbor database, determining a weighted estimated value for the used vehicle nearest neighbor based on the data from the nearest neighbor database, the at least one target used vehicle record, the at least one constraint, and the neighborhood distance function; and

B2) determining an estimated value for the at least one target used vehicle based on the weighted estimated values for the number K of used vehicle nearest neighbors.

3. The method of claim 1, wherein the resale information includes at least one item selected from the group consisting of resale date, region, mileage, condition, and resale price.

4. The method of claim 3, wherein the plurality of vehicle features for each used vehicle nearest neighbor record and the at least one target used vehicle record individually include at least two items selected from the group consisting of vehicle type, model, series, trim level, engine type, transmission type, moon roof equipped, leather interior, interior color, and exterior color.

5. The method of claim 3, wherein the resale information includes resale price and resale region.

6. The method of claim 5, wherein the at least one constraint includes a constraint selected from the group consisting of the pair of vehicles are the same model, the pair of vehicles are the same series, the pair of vehicles have the same model year, the pair of vehicles are the same vehicle series, the difference in mileage between the pair of vehicles is less than about 3,000 miles.

7. The method of claim 6, wherein the at least one used vehicle record further comprises resale information.

8. The method of claim 7, wherein the resale information includes at least one item selected from the group consisting of intended resale date and region.

9. A computer-implemented method for estimating market value of a used vehicle, the method comprising:

A) electronically receiving data from a nearest neighbor database consisting of a number K of used vehicle nearest neighbor records, each used vehicle nearest neighbor record comprising resale information and a plurality of used vehicle features, at least one target used vehicle record comprised of a plurality of used vehicle features, at least one constraint for determining a neighbor relationship between a pair of used vehicles, and a neighborhood distance function for determining a distance between a pair of used vehicles, the number K is iteratively selected for estimation accuracy based on a historical database of N used vehicle records; and

B) electronically determining an estimated value for the at least one target used vehicle based on the data from the nearest neighbor database, the at least one target used vehicle record, the at least one constraint, and the neighborhood distance function,

wherein the determining step includes the use of neural networks, and

wherein the estimated value of the at least one target used vehicle is relied upon by individuals to at least price used vehicles for resale.

10. A computer-implemented method for estimating market value of a used vehicle, the method comprising:

A) electronically receiving data which includes;

V₁ comprised of a number N of v₁, each v₁ comprising resale information and f₁, V₂ comprised of at least one v₂, each v₂ comprised of f₂, Const, F_d, K, and Error_P;

B) determining an Error_K based on V₁, Const, F_d, and K; and

C) if Error_K is less than about Error_P, then

C1) electronically determining an estimated value for each v₂ in V₂ based on V₁, V₂, Const, F_d, and K;

C2) setting K to K plus 1 and Error_P to Error_K; and

C3) looping to step B),

wherein;

V₁ equals data from a historical database comprised of a number N of used vehicle records,

v₁ equals a used vehicle record in V₁,

f₁ equals a plurality of vehicle features of v₁,

V₂ equals a data set comprised of at least one target used vehicle record,

v₂ equals a target used vehicle record,

f₂ equals a plurality of vehicle features of v₂,

Const equals an at least one constraint for determining a neighbor relationship between a pair of used vehicles,

F_d equals a neighborhood distance function for determining a distance between a pair of used vehicles,

K equals a nearest neighbor value,

Error_P equals a previous estimation error, and

Error_K equals a used vehicle market error,

wherein the estimated value of each v₂ in V₂ is relied upon by individuals to at least price used vehicles for resale.

11. The method of claim 10, wherein the resale information includes at least one item selected from the group consisting of resale date, region, mileage, condition, and resale price.

12. The method of claim 10 wherein step B) is comprised of:

B1) for each v_i in V₁,

B11) determining a neighbor group V′

of K used vehicles v′

for v from V₁ based on Const, F_d, and f₁;

B12) for each v′

in V′

, determining a weighted estimated value for v₁ based on v′

, f₁ and F_d;

B13) determining an estimated value for v₁ based on each weighted estimated value of v₁;

B14) determining an estimated error for v₁ based on the estimated value for v₁ and the resale price of v₁; and

B2) determining Error_K based on the estimated error for each v₁ in V₁, and N.

13. The method of claim 10 wherein step C1) is comprised of:

for each v₂ in V₂,

C11) determining a nearest neighbor group V″

of K used vehicles v″

for v₂ from V₁ based on Const, F_d, f₁, and f₂;

C12) for each v″

in V″

, determining a weighted estimated value for v₂ based on v″

, F_d, f₁, and f₂;

C13) determining an estimated value for v₂ based on each weighted estimated values of v₂.

14. The method of claim 13 further comprising C14) storing v₂ with the estimated value for v₂ in a data set V₃ of used vehicles v₃ with estimated market values.

15. The method of claim 10, wherein f₁, and f₂ include at least two items selected from the group consisting of vehicle type, model, series, trim level, engine type, transmission type, moon roof equipped, leather interior, interior color, and exterior color.

16. The method of claim 15, wherein Const includes a constraint selected from the group consisting of the pair of vehicles are the same model, the pair of vehicles are the same series, the pair of vehicles have the same model year, the pair of vehicles are the same vehicle series, the difference in mileage between the pair of vehicles is less than about 3,000 miles.

17. The method of claim 16, wherein each v₂ further comprises planned resale information, wherein the planned resale information includes at least one item selected from the group consisting of intended resale date, region and resale channel.

18. The method of claim 10, wherein the determining step C1) includes the use of neural networks.