System and method for providing a price for a vehicle

Patent metadata

Field	Value
Publication number	US-11004099-B2
Application number	US-201916398933-A
Country	US
Kind code	B2
Filing date	Apr 30, 2019
Priority date	Apr 30, 2019
Publication date	May 11, 2021
Grant date	May 11, 2021

Abstract

System and methods for providing a target price for a target vehicle with a target mileage are provided. One exemplary method includes receiving attributes for the target vehicle, receiving prices for sold vehicles having attributes corresponding to the target vehicle attributes, and receiving mileages for the sold vehicles. The exemplary method may further include generating a linear regression model relating the sold vehicle mileages to the sold vehicle prices, and providing the target price for the target vehicle based on the model.

First claim

What is claimed is:

1. A method for providing a target price for a target vehicle with a target mileage, the method comprising:
receiving, via a graphical user interface, attributes for a target vehicle, the attributes including a target mileage for the target vehicle;
obtaining a plurality of image data corresponding to a plurality of sold vehicles;
creating a training set, the training set including the plurality of image data and a plurality of match scores, the plurality of match scores including at least a first match score between a first image data of the plurality of image data and a second image data of the plurality of image data, a second match score between the second image data and a third image data of the plurality of image data, and a third match score between the first image data and the third image data;
training a machine learning model based on the training set;
obtaining target vehicle image data of the target vehicle;
determining, via the trained machine learning model, a corresponding match score between the target vehicle image data and each of the plurality of image data;
selecting a subset of the plurality of image data from the plurality of image data, such that the corresponding match score between the target vehicle image data and each of the subset of the plurality of image data is above a threshold score;
selecting a subset of sold vehicles from the plurality of sold vehicles based on the selected subset of the plurality of image data;
retrieving sales data for the subset of sold vehicles having attributes corresponding to the attributes of the target vehicle, the sales data including price data of the subset of sold vehicles and mileage data associated with the subset of sold vehicles;
generating a set of non-overlapping mileage ranges, such that one or more mileage values of the mileage data fall within the mileage ranges and the target mileage falls within the mileage ranges;
associating a corresponding bin label with each mileage range of the set of non-overlapping mileage ranges;
selecting a target bin label associated with a mileage range that includes the target mileage; and
generating a target price for the target vehicle by computing an average price of the price data of the subset of sold vehicles associated with the target bin label, wherein generating the target price comprises:
computing an average price of the price data of the subset of sold vehicles associated with each bin label to create a set of bin average prices;
computing bin weights for the bin labels, wherein the bin weights for the bin labels are computed based on a difference between the target mileage and mileage values associated with the bin labels; and
computing a weighted average of bin average prices using the bin weights.

2. The method of claim 1 , wherein a size of any one of the mileage ranges is less than one thousand miles.

3. The method of claim 1 , wherein the bin weights are inversely proportional to a square of the difference between the target mileage and the mileage values associated with the bin labels.

4. The method of claim 1 , further comprising applying a data filter to attributes of the plurality of sold vehicles to create another subset of sold vehicles.

5. The method of claim 4 , wherein applying the data filter comprises filtering the plurality of sold vehicles by a quality score.

6. The method of claim 4 , wherein applying the data filter comprises filtering the plurality of sold vehicles by a match score.

7. The method of claim 4 , wherein applying the data filter comprises filtering the plurality of sold vehicles by one of a year, a make, a model, or a trim line.

8. The method of claim 1 , wherein the attributes of the target vehicle comprise a history of location data for the target vehicle.

9. The method of claim 8 , wherein the attributes of the target vehicle comprise a history of weather data corresponding to the location data history for the target vehicle.

10. The method of claim 1 , wherein the attributes of the subset of sold vehicles comprise a history of location data for the subset of sold vehicles, and the method further comprising applying a data filter to the plurality of sold vehicles by a location of vehicle use.

11. The method of claim 1 , further comprising:
receiving a suggested price for the target vehicle; and
generating a target score based on a difference between the suggested and the target price.

12. The method of claim 1 , further comprising:
obtaining first additional image data, second additional image data, and an additional match score between the first additional image data and the second additional image data;
predicting, via the trained machine learning model, a match score between the first additional image data and the second additional image data;
comparing the predicted match score to the additional match score; and
modifying model parameters of the machine learning model when a difference between the predicted match score and the additional match score is above a threshold value.

13. The method of claim 1 , further comprising:
displaying the graphical user interface, the graphical user interface including a linear regression graph including a plurality of data points corresponding to the price data of the subset of sold vehicles and the mileage data associated with the subset sold vehicles; and
in response to a selection of a data point from the plurality of data points on the linear regression graph, overlaying additional information associated with a corresponding sold vehicle associated with the selected data point over the linear regression graph and in response to another selection of another data point from the plurality of data points on the linear regression graph, overlaying other additional information associated with a corresponding sold vehicle associated with the other selected data point over the linear regression graph.

14. A system for providing a target price for a target vehicle with a target mileage, the system comprising at least one memory storing instructions and at least one processor executing the instructions to perform operations comprising:
receiving, via a graphical user interface, attributes for a target vehicle, the attributes including a target mileage for the target vehicle;
obtaining a plurality of image data corresponding to a plurality of sold vehicles;
creating a training set, the training set including the plurality of image data and a plurality of match scores, the plurality of match scores including at least a first match score between a first image data of the plurality of image data and a second image data of the plurality of image data, a second match score between the second image data and a third image data of the plurality of image data, and a third match score between the first image data and the third image data;
training a machine learning model based on the training set;
obtaining target vehicle image data of the target vehicle;
determining, via the trained machine learning model, a corresponding match score between the target vehicle image data and each of the plurality of image data;
selecting a subset of the plurality of image data from the plurality of image data, such that the corresponding match score between the target vehicle image data and each of the subset of the plurality of image data is above a threshold score;
selecting a subset of sold vehicles from the plurality of sold vehicles based on the selected subset of the plurality of image data;
retrieving sales data for the subset of sold vehicles having attributes corresponding to the attributes of the target vehicle

Assignees

• Capital One Services Llc

Inventors

• Anderson Jacob

Classifications

• G06F30/20

  Design optimisation, verification or simulation (optimisation, verification or simulation of circuit designs G06F30/30) · CPC title

• G06F16/2457

  with adaptation to user needs · CPC title

• G06Q30/0206 Primary

  Price or cost determination based on market factors · CPC title