COMPUTERIZED SYSTEMS, PROCESSES, AND USER INTERFACES FOR TARGETED MARKETING ASSOCIATED WITH A POPULATION OF REAL-ESTATE ASSETS

• US 20150356576A1
• Filed: 05/27/2015
• Published: 12/10/2015
• Est. Priority Date: 05/27/2011
• Status: Abandoned Application

• Current Assignee: Aniruddha Banerjee, Anita Narra, Ashutosh Malaviya, Eric Fang, Fan Jiang, Jason Hiver Tondu, Yu Pan
• Original Assignee: Aniruddha Banerjee, Anita Narra, Ashutosh Malaviya, Eric Fang, Fan Jiang, Jason Hiver Tondu, Yu Pan
• Inventors: Tondu, Jason Hiver, Narra, Anita, Malaviya, Ashutosh, Banerjee, Aniruddha, Pan, Yu, Fang, Eric, Jiang, Fan
• Application Number:
  US14/722,151
• Publication Number:
  US 20150356576A1
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06Q 30/0202 Market predictions or forecasting for commercial activities
• CPC Class Codes: G06Q 50/16 Real estate

Abstract

In one aspect, a method of generating a prediction list of real-estate assets that have a specified probability of being placed for sale within a specified period of time includes the step of providing a list of real-estate assets. Each real-estate asset is associated with one or more real-estate assets attributes. The method includes the step of providing a training data set wherein the training data set comprises a past population of data associated with a plurality of real-estate assets and a set of training-data set attributes for each real-estate asset in the plurality of real-estate assets. The method includes providing a testing data set wherein the testing data set comprises another past population of data associated with the plurality of real-estate assets and a set testing-data set attributes for each real-estate asset in the plurality of real-estate assets, wherein the set of testing data set attributes comprises an updated version of the training data set attributes from a specified later time.

Claims

1. A method of generating a prediction list of real-estate assets that have a specified probability of being placed for sale within a specified period of time comprising:

providing a list of real-estate assets, wherein each real-estate asset is associated with one or more real-estate assets attributes;

providing a training data set wherein the training data set comprises a past population of data associated with a plurality of real-estate assets and a set of training-data set attributes for each real-estate asset in the plurality of real-estate assets;

providing a testing data set wherein the testing data set comprises another past population of data associated with the plurality of real-estate assets and a set testing-data set attributes for each real-estate asset in the plurality of real-estate assets, wherein the set of testing data set attributes comprises an updated version of the training data set attributes from a specified later time;

implementing a backtest on the training data set to determine one or more first prediction models;

generating a first prediction list using the one or more first prediction models, wherein a first probability score for each real-estate asset in the list of real-estate assets to be placed for sale within a specified period of time is calculated using the one or more first prediction models;

using the testing data set to determine a second prediction model from the one or more first prediction models based on the test data set by combining the one or more first prediction models;

generating a second prediction list using the second prediction model, wherein a second probability score for each real-estate asset in the list of real-estate assets to be placed for sale within the specified period of time is calculated using the second prediction model;

averaging the first probability score and the second probability score of each real-estate asset in the list of real-estate assets to generate an averaged probability score for each real-estate asset; and

ordering a prediction list comprising each real-estate asset ordered according for each real-estate asset'"'"'s averaged probability score.

2. The method of claim 1, wherein a real-estate assets comprises a residential real-estate home.

3. The method of claim 1, wherein the one or more first prediction models comprise two champion logistic-properties prediction models.

4. The method of claim 3, wherein the one or more first prediction models comprise a balanced-random-forest model.

5. The method of claim 4, wherein the one or more first prediction models comprises an unbalanced-random-forest prediction model.

6. The method of claim 5, wherein the testing data set is used to tune the weights of the one or more first prediction models.

7. The method of claim 1, wherein the training data set comprises a two-years previous past population of data.

8. The method of claim 1, wherein the testing data set comprises a one-year previous past population of data.

9. A computerized system generating a prediction list of real-estate assets that have a specified probability of being placed for sale within a specified period of time comprising:

a processor configured to execute instructions;

a memory containing instructions when executed on the processor, causes the processor to perform operations that;

provide a list of real-estate assets, wherein each real-estate asset is associated with one or more real-estate assets attributes;

provide a training data set wherein the training data set comprises a past population of data associated with a plurality of real-estate assets and a set of training-data set attributes for each real-estate asset in the plurality of real-estate assets;

provide a testing data set wherein the testing data set comprises another past population of data associated with the plurality of real-estate assets and a set testing-data set attributes for each real-estate asset in the plurality of real-estate assets, wherein the set of testing data set attributes comprises an updated version of the training data set attributes from a specified later time;

implement a backtest on the training data set to determine one or more first prediction models;

generate a first prediction list using the one or more first prediction models, wherein a first probability score for each real-estate asset in the list of real-estate assets to be placed for sale within a specified period of time is calculated using the one or more first prediction models;

use the testing data set to determine a second prediction model from the one or more first prediction models based on the test data set by combining the one or more first prediction models;

generate a second prediction list using the second prediction model, wherein a second probability score for each real-estate asset in the list of real-estate assets to be placed for sale within the specified period of time is calculated using the second prediction model;

average the first probability score and the second probability score of each real-estate asset in the list of real-estate assets to generate an averaged probability score for each real-estate asset; and

order a prediction list comprising each real-estate asset ordered according for each real-estate asset'"'"'s averaged probability score.

10. The computerized system of claim 9, wherein a real-estate assets comprises a residential real-estate home.

11. The computerized system of claim 10, wherein the one or more first prediction models comprise two champion logistic-properties prediction models.

12. The computerized system of claim 11, wherein the one or more first prediction models comprise a balanced-random-forest model.

13. The computerized system of claim 12, wherein the one or more first prediction models comprises an unbalanced-random-forest prediction model.

14. The computerized system of claim 13, wherein the testing data set is used to tune the weights of the one or more first prediction models.

15. The computerized system of claim 14, wherein the training data set comprises a two-years previous past population of data.

16. The computerized system of claim 15, wherein the testing data set comprises a one-year previous past population of data.