IP Library › Granted Patent US 12670436
Granted Patent B2
US 12670436 · App. 17/957,127 · Granted Jun 30, 2026

Systems and methods for predicting differentiating features

Inventors: Aamer Charania (Flower Mound, TX); Abhisek Jana (Herndon, VA); Jiankun Liu (Flower Mound, TX); Behrouz Saghafi Khadem (Frisco, TX)
Assignee: Capital One Services, LLC
G06N20/00G06F18/21375
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12670436
App. No.
17/957,127
Granted
Jun 30, 2026
Kind
B2
Abstract

Systems and methods for predicting differentiating features from tabular data for two different populations. In some aspects, the systems and methods provide for receiving first data entries and second data entries and generating a first graph based on a data entry from the first data entries and a second graph based on a data entry from the second data entries. A first node in the first graph is determined to correspond to a second node in the second graph. A first set of graph embeddings is generated based on the first graph. A second set of graph embeddings is generated based on the second graph. Using a machine learning model, the first set of graph embeddings and the second set of graph embeddings are processed to identify at least one feature indicative of a difference between the first data entries and the second data entries.

Claims (45)

1 . A system for predicting one or more features indicative of a difference between a first population and a second population, the system comprising:

one or more processors; and

a non-transitory, computer-readable medium comprising instructions that, when executed by the one or more processors, cause operations comprising:

receiving in tabular form a plurality of first data entries for a first population and a plurality of second data entries for a second population, wherein:

a first data entry of the plurality of first data entries comprises first values for one or more features; and

a second data entry of the plurality of second data entries comprises second values for the one or more features;

generating a first graph based on the first data entry and a second graph based on the second data entry, wherein:

the first graph is independent of any events before or after a first time stamp; and

the second graph is independent of any events before or after a second time stamp;

determining that a first node in the first graph associated with the first time stamp corresponds to a second node in the second graph associated with the second time stamp;

generating, based on determining that the first node corresponds to the second node and for the first graph, a first set of graph embeddings that capture information regarding one or more of a graph topology of the first graph, node-to-node relationship of the first graph, or other information regarding subgraphs or nodes of the first graph;

generating, based on determining that the first node corresponds to the second node and for the second graph, a second set of graph embeddings that capture information regarding one or more of a graph topology of the second graph, node-to-node relationship of the second graph, or other information regarding subgraphs or nodes of the second graph; and

processing, using a machine learning model, the first set of graph embeddings and the second set of graph embeddings to identify, from the one or more features, at least one feature with a probability satisfying a threshold to be indicative of a difference between the first population and the second population.

2 . The system of claim 1 , wherein the first population is representative of a population at a first time, and wherein the second population is representative of the population at a second time.

3 . The system of claim 1 , wherein the first population is related to a first geographical location, and wherein the second population is related to a second geographical location.

4 . A method comprising:

receiving a plurality of first data entries and a plurality of second data entries;

generating a first graph based on a first data entry of the plurality of first data entries and a second graph based on a second data entry of the plurality of second data entries;

determining that a first node in the first graph corresponds to a second node in the second graph;

generating, based on determining that the first node corresponds to the second node and for the first graph, a first set of graph embeddings that capture information regarding one or more of a graph topology of the first graph, node-to-node relationship of the first graph, or other information regarding subgraphs or nodes of the first graph;

generating, based on determining that the first node corresponds to the second node and for the second graph, a second set of graph embeddings that capture information regarding one or more of a graph topology of the second graph, node-to-node relationship of the second graph, or other information regarding subgraphs or nodes of the second graph; and

processing, using a machine learning model, the first set of graph embeddings and the second set of graph embeddings to identify at least one feature common to the first data entry and the second data entry and indicative of a difference between the plurality of first data entries and the plurality of second data entries.

5 . The method of claim 4 , wherein the plurality of first data entries corresponds to a first population, and wherein the plurality of second data entries corresponds to a second population.

6 . The method of claim 5 , wherein the first population is representative of a population at a first time, and wherein the second population is representative of the population at a second time.

7 . The method of claim 5 , wherein the first population is related to a first geographical location, and wherein the second population is related to a second geographical location.

8 . The method of claim 4 , wherein the plurality of first data entries is representative of source data used to train a model, and wherein the plurality of second data entries is representative of predictive data generated using the model.

9 . The method of claim 4 , wherein the machine learning model comprises an Euclidean distance based model, a naive Bayesian model, or an encoder-decoder model.

10 . The method of claim 4 , wherein the first data entry comprises first values for one or more features, and wherein the second data entry comprises second values for the one or more features.

11 . The method of claim 10 , wherein the first graph is independent of any events before or after a first time stamp, wherein the second graph is independent of any events before or after a second time stamp, and wherein determining that a first node in the first graph corresponds to a second node in the second graph comprises determining that the first node in the first graph associated with the first time stamp corresponds to the second node in the second graph associated with the second time stamp.

12 . The method of claim 11 , wherein identifying the at least one feature comprises identifying the at least one feature from the one or more features.

13 . A non-transitory, computer-readable medium comprising instructions that, when executed by one or more processors, cause operations comprising:

receiving a plurality of first data entries and a plurality of second data entries;

generating a first graph based on a first data entry of the plurality of first data entries and a second graph based on a second data entry of the plurality of second data entries;

determining that a first node in the first graph corresponds to a second node in the second graph;

generating, based on determining that the first node corresponds to the second node and for the first graph, a first set of graph embeddings that capture information regarding one or more of a graph topology of the first graph, node-to-node relationship of the first graph, or other information regarding subgraphs or nodes of the first graph;

generating, based on determining that the first node corresponds to the second node and for the second graph, a second set of graph embeddings that capture information regarding one or more of a graph topology of the second graph, node-to-node relationship of the second graph, or other information regarding subgraphs or nodes of the second graph; and

processing, using a machine learning model, the first set of graph embeddings and the second set of graph embeddings to identify at least one feature common to the first data entry and the second data entry and indicative of a difference between the plurality of first data entries and the plurality of second data entries.

14 . The non-transitory, computer-readable medium of claim 13 , wherein the plurality of first data entries corresponds to a first population, and wherein the plurality of second data entries corresponds to a second population.

15 . The non-transitory, computer-readable medium of claim 14 , wherein the first population is representative of a population at a first time, and wherein the second population is representative of the population at a second time.

16 . The non-transitory, computer-readable medium of claim 14 , wherein the first population is related to a first geographical location, and wherein the second population is related to a second geographical location.

17 . The non-transitory, computer-readable medium of claim 13 , wherein the plurality of first data entries is representative of source data used to train a model, and wherein the plurality of second data entries is representative of predictive data generated using the model.

18 . The non-transitory, computer-readable medium of claim 13 , wherein the machine learning model comprises an Euclidean distance based model, a naive Bayesian model, or an encoder-decoder model.

19 . The non-transitory, computer-readable medium of claim 13 , wherein the first data entry comprises first values for one or more features, and wherein the second data entry comprises second values for the one or more features.

20 . The non-transitory, computer-readable medium of claim 19 , the first graph is independent of any events before or after a first time stamp, wherein the second graph is independent of any events before or after a second time stamp, and wherein determining that a first node in the first graph corresponds to a second node in the second graph comprises determining that the first node in the first graph associated with the first time stamp corresponds to the second node in the second graph associated with the second time stamp.

21 . The non-transitory, computer-readable medium of claim 20 , wherein identifying the at least one feature comprises identifying the at least one feature from the one or more features.