Granted Patent B2

US 12,387,143 · App. 18/516,673 · Granted Aug 12, 2025

Inventors: Suresh B. Gajendran (Richardson, TX); Mahesh Chandrappa (Bloomington, IL); Mark A. Dickneite (Bloomington, IL); Charles T. Fiala (Normal, IL); Rashid Zaheer (Bloomington, IL)

Assignee: State Farm Mutual Automobile Insurance Company

G06N20/00G06F8/34G06F9/453G06F18/214G06F40/35G06N3/045

Abstract

Claims (105)

1. A method implemented by a computing device, the method comprising:

generating a guided user interface (GUI) configured to enable construction of machine learning (ML) model generation tools;

receiving inputs, via the GUI, indicative of requested operations to be performed by the computing device, the requested operations including:

obtaining, from a database, a dataset associated with a plurality of data objects,

generating, on the GUI, a visualization indicating a correlation between at least two objects of the plurality of data objects,

determining, based on the visualization, a subset of the dataset correlated with a first data object,

selecting a ML algorithm, and

generating, based on the subset of the dataset and using the ML algorithm, an ML model generation tool corresponding to the first data object; and

implementing the ML model generation tool in a cloud server.

2. The method of claim 1 , wherein the requested operations further comprise:

generating, using the ML model generation tool, an ML model corresponding to the first data object; and

implementing the ML model, using the first data object, in the cloud server.

3. The method of claim 2 , further comprising:

receiving a request to predict a target value associated with a target object, the request including a new dataset;

determining whether an ML model with respect to the target object exists in the cloud server; and

in response to the determination that the ML model with respect to the target object exists in the cloud server,

downloading the ML model with respect to the target object from the cloud server to a local storage, and

computing, using the ML model with respect to the target object, the target value.

4. The method of claim 3 , further comprising:

in response to the determination that the ML model with respect to the target object does not exist in the cloud server,

downloading the ML model generation tool from the cloud server to the local storage;

executing the ML model generation tool to generate, based at least in part on the new dataset, the ML model with respect to the target object; and

compute, using the ML model with respect to the target object, the target value.

5. The method of claim 1 , wherein the requested operations further comprise:

generating, on the GUI, a first visualization of linear dependencies between the first data object and other data objects of the plurality of data objects;

determining, based on the first visualization, one or more second data objects having the linear dependencies higher than a threshold; and

performing, based at least in part on the one or more second data objects, a dimension reduction on the dataset to obtain the subset of the dataset.

6. The method of claim 5 , wherein the dimension reduction is performed using at least one of a random forest algorithm, a single variable logistic regression algorithm, or a variable clustering algorithm.

7. The method of claim 1 , wherein the requested operations further comprise:

causing the dataset to be transmitted from a database to a local storage;

presenting, on the GUI, statistic information associated with the dataset; and

performing at least one of a null value treatment or an outlier value treatment on the dataset.

8. A system comprising:

a processor; and

memory storing instructions that, when executed by processor, cause the processor to perform actions comprising:

generating a guided user interface (GUI) configured to enable construction of machine learning (ML) model generation tools;

receiving inputs, via the GUI, indicative of requested operations to be performed by the computing device, the requested operations including:

obtaining, from a database, a dataset associated with a plurality of data objects,

generating, on the GUI, a visualization indicating a correlation between at least two objects of the plurality of data objects,

determining, based on the visualization, a subset of the dataset correlated with a first data object,

selecting a ML algorithm, and

generating, based on the subset of the dataset and using the ML algorithm, an ML model generation tool corresponding to the first data object; and

implementing the ML model generation tool in a cloud server.

9. The system of claim 8 , wherein the requested operations further comprise:

generating, using the ML model generation tool, an ML model corresponding to the first data object; and

implementing the ML model, using the first data object, in the cloud server.

10. The system of claim 9 , wherein the processor is caused to further perform actions including:

receiving a request to predict a target value associated with a target object, the request including a new dataset;

determining whether an ML model with respect to the target object exists in the cloud server; and

in response to the determination that the ML model with respect to the target object exists in the cloud server,

downloading the ML model with respect to the target object from the cloud server to a local storage, and

computing, using the ML model with respect to the target object, the target value.

11. The system of claim 10 , wherein the processor is caused to further perform actions including:

in response to the determination that the ML model with respect to the target object does not exist in the cloud server,

downloading the ML model generation tool from the cloud server to the local storage;

executing the ML model generation tool to generate, based at least in part on the new dataset, the ML model with respect to the target object; and

compute, using the ML model with respect to the target object, the target value.

12. The system of claim 8 , wherein the requested operations further comprise:

generating, on the GUI, a first visualization of linear dependencies between the first data object and other data objects of the plurality of data objects;

determining, based on the first visualization, one or more second data objects having the linear dependencies higher than a threshold; and

performing, based at least in part on the one or more second data objects, a dimension reduction on the dataset to obtain the subset of the dataset.

13. The system of claim 12 , wherein the dimension reduction is performed using at least one of a random forest algorithm, a single variable logistic regression algorithm, or a variable clustering algorithm.

14. The system of claim 8 , wherein the requested operations further comprise:

causing the dataset to be transmitted from a database to a local storage;

presenting, on the GUI, statistic information associated with the dataset; and

performing at least one of a null value treatment or an outlier value treatment on the dataset.

15. A computer-readable storage medium storing computer-readable instructions executable by a processor, that when executed by the processor, cause the processor to perform actions comprising:

generating a guided user interface (GUI) configured to enable construction of machine learning (ML) model generation tools;

receiving inputs, via the GUI, indicative of requested operations to be performed by the computing device, the operations including:

obtaining, from a database, a dataset associated with a plurality of data objects,

generating, on the GUI, a visualization indicating a correlation between at least two objects of the plurality of data objects,

determining, based on the visualization, a subset of the dataset correlated with a first data object,

selecting a ML algorithm, and

generating, based on the subset of the dataset and using the ML algorithm, an ML model generation tool corresponding to the first data object; and

implementing the ML model generation tool in a cloud server.

16. The computer-readable storage medium of claim 15 , wherein the requested operations further comprise:

generating, on the GUI, a first visualization of linear dependencies between the first data object and other data objects of the plurality of data objects;

determining, based on the first visualization, one or more second data objects having the linear dependencies higher than a threshold; and

performing, based at least in part on the one or more second data objects, a dimension reduction on the dataset to obtain the subset of the dataset.

17. The computer-readable storage medium of claim 16 , wherein the requested operations further comprise:

causing the dataset to be transmitted from a database to a local storage;

presenting, on the GUI, statistic information associated with the dataset; and

performing at least one of a null value treatment or an outlier value treatment on the dataset.

18. The computer-readable storage medium of claim 15 , wherein the processor is caused to further perform actions including:

receiving a request to predict a target value associated with a target object, the request including a new dataset;

determining whether an ML model with respect to the target object exists in the cloud server;

in response to the determination that the ML model with respect to the target object exists in the cloud server,

downloading the ML model with respect to the target object from the cloud server to a local storage, and

computing, using the ML model with respect to the target object, the target value; and

in response to the determination that the ML model with respect to the target object does not exist in the cloud server,

downloading the ML model generation tool from the cloud server to the local storage;

executing the ML model generation tool to generate, based at least in part on the new dataset, the ML model with respect to the target object; and

compute, using the ML model with respect to the target object, the target value.

19. The computer-readable storage medium of claim 15 , wherein the requested operations further comprise:

generating, using the ML model generation tool, an ML model corresponding to the first data object; and

implementing the ML model, using the first data object, in the cloud server.

20. A system comprising:

means for generating a guided user interface (GUI) configured to enable construction of machine learning (ML) model generation tools;

means for receiving inputs, via the GUI, indicative of requested operations to be performed by the computing device, the operations including:

means for obtaining, from a database, a dataset associated with a plurality of data objects,

means for generating, on the GUI, a visualization indicating a correlation between at least two objects of the plurality of data objects,

means for determining, based on the visualization, a subset of the dataset correlated with a first data object,

means for selecting a ML algorithm, and

means for generating, based on the subset of the dataset and using the ML algorithm, an ML model generation tool corresponding to the first data object; and

means for implementing the ML model generation tool in a cloud server.

Assignments (1)

Continuity (3)

References Cited (23)