Granted Patent B2

US 12,363,021 · App. 18/205,945 · Granted Jul 15, 2025

Inventor: Sebastian Maurice (Toronto, CA)

Assignee: Ayla Networks, Inc.

H04L43/55H04L41/0661H04L41/0677H04L41/083H04L41/0836H04W24/06

Abstract

Claims (58)

1. A method comprising:

receiving, from one or more subsystems, one or more predicted outcomes associated with a device connected to a gateway device, wherein each of the one or more predicted outcomes is associated with at least one feature of the device;

providing a subset of the one or more predicted outcomes as input to a machine learning model trained to identify a set of resolution actions for the subset of the one or more predicted outcomes;

receiving, from the machine learning model, the set of resolution actions for the subset of the one or more predicted outcomes, wherein each resolution action in the set of resolution actions is associated with a probability of resolving at least one of the predicted outcomes in the subset of the one or more predicted outcomes;

identifying a first resolution action from the set of resolution actions, wherein the first resolution action has a highest probability of resolving the at least one of the predicted outcomes in the subset of the one or more predicted outcomes;

identifying using a multiple linear regression algorithm, a first set of features negatively correlated with the subset of the one or more predicted outcomes, wherein the first set of features comprises a first subset of gateway features associated with the gateway device and a first subset of device features associated with the device, and wherein the first set of features comprises features that have a first test value associated with the multiple linear regression algorithm that satisfies a first threshold;

identifying, using the multiple linear regression algorithm, a second set of features positively correlated with the subset of the one or more predicted outcomes, wherein the second set of features comprises a second subset of the gateway features associated with the gateway device and a second subset of the device features associated with the device, and wherein the second set of features comprises features that have a second test value associated with the multiple linear regression algorithm that satisfies a second threshold;

identifying, based on the first set of features negatively correlated with the subset of the one or more predicted outcomes and the second set of features positively correlated with the subset of the one or more predicted outcomes, a root cause associated with the first resolution action; and

providing, to the device, a first instruction to execute the first resolution action.

2. The method of claim 1 , wherein the first set of features and the second set of features are identified using a correlation matrix.

3. The method of claim 2 , wherein the first set of features negatively correlated with the subset of the one or more predicted outcomes comprises features that have a first correlation value that satisfies a first correlation threshold, and wherein the second set of features positively correlated with the subset of the one or more predicted outcomes comprises features that have a second correlation value that satisfies a second correlation threshold.

4. The method of claim 2 , further comprising:

identifying, using the correlation matrix, the first set of features negatively correlated with the subset of the one or more predicted outcomes for a predetermined period of time preceding a time at which the first resolution action from the set of resolution actions is identified; and

identifying, using the correlation matrix, the second set of features positively correlated with the subset of the one or more predicted outcomes for the predetermined period of time preceding the time at which the first resolution action from the set of resolution actions is identified.

5. The method of claim 1 , further comprising:

receiving a training dataset comprising a plurality of predicted outcomes and a plurality of resolution actions;

providing the training dataset as input into an untrained machine learning model; and

training the untrained machine learning model based on the training dataset to generate the machine learning model trained to identify the set of resolution actions.

6. The method of claim 1 , further comprising:

providing, to the device, at least one of the first resolution action, the at least one of the predicted outcomes, or an indication of the root cause associated with the first resolution action.

7. A system comprising:

a memory; and

a processing device operatively coupled to the memory, the processing device to:

receive, from one or more subsystems, one or more predicted outcomes associated with a device connected to a gateway device, wherein each of the one or more predicted outcomes is associated with at least one feature of the device;

provide a subset of the one or more predicted outcomes as input to a machine learning model trained to identify a set of resolution actions for the subset of the one or more predicted outcomes;

receive, from the machine learning model, the set of resolution actions for the subset of the one or more predicted outcomes, wherein each resolution action in the set of resolution actions is associated with a probability of resolving at least one of the predicted outcomes in the subset of the one or more predicted outcomes;

identify a first resolution action from the set of resolution actions, wherein the first resolution action has a highest probability of resolving the at least one of the predicted outcomes in the subset of the one or more predicted outcomes;

identify using a multiple linear regression algorithm, a first set of features negatively correlated with the subset of the one or more predicted outcomes, wherein the first set of features comprises a first subset of gateway features associated with the gateway device and a first subset of device features associated with the device, and wherein the first set of features comprises features that have a first test value associated with the multiple linear regression algorithm that satisfies a first threshold;

identify, using the multiple linear regression algorithm, a second set of features positively correlated with the subset of the one or more predicted outcomes, wherein the second set of features comprises a second subset of the gateway features associated with the gateway device and a second subset of the device features associated with the device, and wherein the second set of features comprises features that have a second test value associated with the multiple linear regression algorithm that satisfies a second threshold;

identify, based on the first set of features negatively correlated with the subset of the one or more predicted outcomes ad the second set of features positively correlated with the subset of the one or more predicted outcomes, a root cause associated with the first resolution action; and

provide, to the device, a first instruction to execute the first resolution action.

8. The system of claim 7 , wherein the first set of features and the second set of features are identified using a correlation matrix.

9. The system of claim 8 , wherein the first set of features negatively correlated with the subset of the one or more predicted outcomes comprises features that have a first correlation value that satisfies a first correlation threshold, and wherein the second set of features positively correlated with the subset of the one or more predicted outcomes comprises features that have a second correlation value that satisfies a second correlation threshold.

10. The system of claim 8 , wherein the processing device is to:

identify, using the correlation matrix, the first set of features negatively correlated with the subset of the one or more predicted outcomes for a predetermined period of time preceding a time at which the first resolution action from the set of resolution actions is identified; and

identify, using the correlation matrix, the second set of features positively correlated with the subset of the one or more predicted outcomes for the predetermined period of time preceding the time at which the First resolution action from the set of resolution actions is identified.

11. The system of claim 7 , wherein the processing device is to:

provide, to the device, at least one of the first resolution action, the at least one of the predicted outcomes, or an indication of the root cause associated with the first resolution action.

12. A non-transitory computer readable storage medium comprising instructions that, when executed by a processing device, cause the processing device to perform operations comprising:

receiving, from one or more subsystems, one or more predicted outcomes associated with a device connected to a gateway device, wherein each of the one or more predicted outcomes is associated with at least one feature of the device;

providing a subset of the one or more predicted outcomes as input to a machine learning model trained to identify a set of resolution actions for the subset of the one or more predicted outcomes;

receiving, from the machine learning model, the set of resolution actions for the subset of the one or more predicted outcomes, wherein each resolution action in the set of resolution actions is associated with a probability of resolving at least one of the predicted outcomes in the subset of the one or more predicted outcomes;

identifying a first resolution action from the set of resolution actions, wherein the first resolution action has a highest probability of resolving the at least one of the predicted outcomes in the subset of the one or more predicted outcomes;

identifying, using a multiple linear regression algorithm, a first set of features negatively correlated with the subset of the one or more predicted outcomes, wherein the first set of features comprises a first subset of gateway features associated with the gateway device and a first subset of device features associated with the device and wherein the first set of features comprises features that have a first test value associated with the multiple linear regression algorithm that satisfies a first threshold;

identifying, using the multiple linear regression algorithm, a second set of features positively correlated with the subset of the one or more predicted outcomes, wherein the second set of features comprise a second subset of the gateway features associated with the gateway device and a second subset of the device features associated with the device, and wherein the second set of features comprises features that have a second test value associated with the multiple linear regression algorithm that satisfies a second threshold;

identifying, based on the first set of features negatively correlated with the subset of the one or more predicted outcomes and the second set of features positively correlated with the subset of the one or more predicted outcomes, a root cause associated with the first resolution action; and

providing, to the device, a first instruction to execute the first resolution action.

13. The non-transitory computer readable storage medium of claim 12 , wherein the first set of features and the second set of features are identified using a correlation matrix.

14. The non-transitory computer readable storage medium of claim 13 , wherein the first set of features negatively correlated with the subset of the one or more predicted outcomes comprises features that have a first correlation value that satisfies a first correlation threshold, and wherein the second set of features positively correlated with the subset of the one or more predicted outcomes comprises features that have a second correlation value that satisfies a second correlation threshold.

15. The non-transitory computer readable storage medium of claim 13 , the operations further comprising:

identifying, using the correlation matrix, the first set of features negatively correlated with the subset of the one or more predicted outcomes for a predetermined period of time preceding a time at which the first resolution action from the set of resolution actions is identified; and

identifying, using the correlation matrix, the second set of features positively correlated with the subset of the one or more predicted outcomes for the predetermined period of time preceding the time at which the first resolution action from the set of resolution actions is identified.

16. The non-transitory computer readable storage medium of claim 13 , the operations further comprising:

receiving a training dataset comprising a plurality of predicted outcomes and a plurality of resolution actions;

providing the training dataset as input into an untrained machine learning model; and

training the untrained machine learning model based on the training dataset to generate the machine learning model trained to identify the set of resolution actions.

17. The non-transitory computer readable storage medium of claim 13 , the operations further comprising:

providing, to the device, at least one of the first resolution action, the at least one of the predicted outcomes, or an indication of the root cause associated with the first resolution action.

Assignments (2)

Continuity (2)

References Cited (34)