Systems and methods for using machine learning models to automatically identify and compensate for recurring charges

Patent metadata

Field	Value
Publication number	US-12136122-B2
Application number	US-202217947216-A
Country	US
Kind code	B2
Filing date	Sep 19, 2022
Priority date	Oct 19, 2020
Publication date	Nov 5, 2024
Grant date	Nov 5, 2024

Abstract

Disclosed embodiments may include a method and system for automated incremental payments. The system may identify recurring charges from historical account data. Based on the recurring charges and an incremental period, the system may calculate an incremental amount and expected amount. At each iteration of the incremental period, the incremental amount may be assigned to a savings bucket. The value of the savings bucket may be subtracted from an actual account balance to calculate a reduced account balance. The system may generate and transmit a graphical user interface to a user device showing the reduced account balance. The system may receive current data containing a charge that corresponds to the recurring charges. The system may reduce the value of the savings bucket by the amount of the current data charge. If the current data charge is different from the expected amount, the system may adjust the incremental amount accordingly.

First claim

What is claimed is:

1. An incremental data designation system comprising:
one or more processors; and
memory in communication with the one or more processors and storing instructions that, when executed by the one or more processors, are configured to cause the system to:
receive historical data associated with a user;
train, using training data comprising historical data associated with other users, a first machine learning model to predict charges for the user;
identify, using the first machine learning model, a set of recurring charges from the historical data associated with the user;
monitor, by the first machine learning model, an account associated with the user to update the set of recurring charges in response to identifying new or changing recurring charges;
calculate an incremental amount based on an incremental period and the set of recurring charges, wherein the incremental period is a predetermined period of time comprising a predetermined number of discrete time increments and the incremental amount multiplied by the incremental period is an expected amount;
assign, at each successive time increment of the predetermined number of discrete time increments of the incremental period, the incremental amount to a savings bucket comprising a value;
generate a graphical user interface for displaying a reduced balance equal to an actual balance minus the value of the savings bucket;
transmit the graphical user interface to a user device;
receive current data;
extract a second amount from the current data;
determine, using a second machine learning model, that the second amount corresponds to the set of recurring charges; and
reduce the value of the savings bucket by the second amount.

2. The incremental data designation system of claim 1 , wherein the memory comprises storing instructions that are further configured to cause the system to:
determine that the second amount is different from the expected amount; and
adjust, using a third machine learning model, the incremental amount based on the second amount.

3. The incremental data designation system of claim 1 , wherein the memory comprises storing instructions that are further configured to cause the system to:
transmit, to the user device, a prompt for the user to input the set of recurring charges and the incremental period; and
receive, from the user device, an input of the set of recurring charges and the incremental period.

4. The incremental data designation system of claim 1 , wherein identifying, with the first machine learning model, the set of recurring charges from the historical data is based on a location, purchases reoccurring from a same source, purchases at a same time of year, or combinations thereof.

5. The incremental data designation system of claim 1 , wherein the first machine learning model and the second machine learning model are a long short-term memory neural network or an ensemble of stateful long short-term memory neural networks.

6. The incremental data designation system of claim 1 , wherein the historical data comprises a plurality of charges each comprising a source name, an amount, a time, and a location.

7. The incremental data designation system of claim 1 , wherein the memory comprises storing instructions that are further configured to cause the system to:
dynamically modify the graphical user interface to display a modified reduced balance at each iteration of the incremental period.

8. An incremental data designation system comprising:
one or more processors; and
memory in communication with the one or more processors and storing instructions that, when executed by the one or more processors, are configured to cause the system to:
receive historical data associated with a user;
train, using training data comprising historical data associated with other users, a first machine learning model to predict charges for the user;
identify, using the first machine learning model, a set of recurring charges from the historical data associated with the user, wherein the first machine learning model is trained;
monitor, by the first machine learning model, an account associated with the user to update the set of recurring charges in response to identifying new or changing recurring charges;
calculate a first incremental amount based on an incremental period and the set of recurring charges, wherein the first incremental amount multiplied by the incremental period is an expected amount and the incremental period is a predetermined period of time;
assign, at each successive time increment of a predetermined number of discrete time increments of the incremental period, the first incremental amount to a savings bucket comprising a value;
calculate a reduced balance equal to an actual balance minus the value of the savings bucket;
receive new data;
extract a second amount from the new data;
determine, using a second machine learning model, that the second amount corresponds to the set of recurring charges;
determine whether the second amount is different from the expected amount;
responsive to determining that the second amount is different from the expected amount:
adjust the first incremental amount based on the second amount; and
change the value of the savings bucket and the reduced balance based on the second amount and the expected amount.

9. The incremental data designation system of claim 8 , wherein the memory comprises storing instructions that are further configured to cause the system to:
generate a first graphical user interface for displaying the reduced balance; and
transmit the first graphical user interface to a user device.

10. The incremental data designation system of claim 8 , wherein determining, with the second machine learning model, that the second amount corresponds to the set of recurring charges is based on a location, purchases reoccurring from a same source or purchases at a same time of year.

11. The incremental data designation system of claim 8 , wherein adjusting the first incremental amount uses a third machine learning model, wherein the first incremental amount is adjusted based on periodic cycles that affect balance trends.

12. The incremental data designation system of claim 11 , wherein the memory comprises storing instructions that are further configured to cause the system to:
calculate that the expected amount does not cover periodic cycle balance trends of the user; and
responsive to calculating that the expected amount does not cover the periodic cycle balance trends of the user:
generate a second graphical user interface comprising a warning that the first incremental amount is too low; and
transmit the second graphical user interface to a user device.

13. The incremental data designation system of claim 8 , wherein changing the value of the savings bucket based on the second amount and the expected amount further comprises:
determining whether the second amount is more than the expected amount;
responsive to determining that the second amount is more than the expected amount:
reducing the value of the savings bucket by the expected amount;
reducing the reduced balance by a first difference of the second amount and the expected amount;

responsive to determining that the second amount is not more than the expected amount:
reducing the value of the savings bucket by the second amount; and
increasing the reduced balance by a second difference of the expected amount and the second amount.

14. The incremental data designation system of claim 8 , wherein the memory comprises storing instructions that are further configured to cause the sy

Assignees

• Capital One Services Llc

Inventors

• Wallace Christopher
• Eden Grant
• Barr Brian
• Sharpe Samuel
• Truong Anh
• Walters Austin

Classifications

• G06F16/2453

  Query optimisation · CPC title

• G06F16/211

  Schema design and management · CPC title

• G06F16/2282 Primary

  Tablespace storage structures; Management thereof · CPC title

• G06N3/08

  Learning methods · CPC title

• G06N3/044

  Recurrent networks, e.g. Hopfield networks · CPC title