Network-based machine learning microservice platform

Patent metadata

Field	Value
Publication number	US-11544617-B2
Application number	US-201815960265-A
Country	US
Kind code	B2
Filing date	Apr 23, 2018
Priority date	Apr 23, 2018
Publication date	Jan 3, 2023
Grant date	Jan 3, 2023

Abstract

A method may include a processing system having at least one processor for receiving a first machine learning model, the first machine learning model in a first format associated with a first development environment, adapting the first machine learning model to a containerized environment, validating the first machine learning model according to at least one validation criterion associated with a repository, and publishing the first machine learning model to the repository.

First claim

What is claimed is:

1. A method comprising:
receiving, by a processing system including at least one processor, a first machine learning model, wherein the first machine learning model is in a first format associated with a first development environment;
adapting, by the processing system, the first machine learning model to a containerized environment, wherein the adapting the first machine learning model to the containerized environment comprises:
generating a first artifact associated with the first machine learning model, wherein the first artifact includes library information for deployment of the first machine learning model in the containerized environment; and
defining at least a second artifact associated with the first machine learning model, wherein the at least the second artifact defines an external compatibility of at least one of: an input port or an output port of the first machine learning model;

validating, by the processing system, the first machine learning model according to at least one validation criterion associated with a repository; and
publishing, by the processing system, the first machine learning model to the repository.

2. The method of claim 1 , wherein the at least one validation criterion comprises:
an outcome of an application of a test data set to the first machine learning model in a simulation.

3. The method of claim 1 , wherein the validating the first machine learning model according to the at least one validation criterion associated with the repository comprises:
validating the first artifact and the second artifact associated with the first machine learning model.

4. The method of claim 1 , further comprising:
training the first machine learning model with a training data set.

5. The method of claim 1 , further comprising:
validating a composite solution including the first machine learning model according to the at least one validation criterion associated with the repository.

6. The method of claim 5 , wherein the validating the composite solution comprises:
determining a compatibility of at least the second artifact associated with the first machine learning model with at least a third artifact, wherein the first machine learning model comprises a first process of the composite solution, wherein the at least the third artifact is associated with a second process of the composite solution that is stored in the repository and is compatible with the containerized environment.

7. The method of claim 6 , wherein the second artifact defines the compatibility of the at least one of the input port or the output port of the first machine learning model with at least one of an input port or an output port of the second process.

8. The method of claim 5 , further comprising:
training the composite solution with a training data set.

9. The method of claim 5 , wherein the composite solution is received from a user workstation.

10. The method of claim 5 , further comprising:
deploying the composite solution to process a data stream in a network.

11. The method of claim 5 , further comprising:
publishing the composite solution to the repository.

12. The method of claim 1 , wherein the first machine learning model is published to the repository as a microservice.

13. The method of claim 12 , wherein the microservice comprises an executable package comprising a set of artifacts to enable a performance of a data processing task, the set of artifacts including:
at least one script defining the first machine learning model;
at least a first library associated with the first development environment;
configuration data for the first machine learning model; and
the second artifact.

14. The method of claim 13 , wherein the set of artifacts further includes:
the first artifact.

15. The method of claim 1 , wherein the repository includes a search function for searching for at least one microservice stored in the repository based upon at least one of:
a topic;
a popularity;
a type of function;
a ranking based upon at least one performance metric; or
an author.

16. The method of claim 15 , wherein each of the at least one microservice comprises one of:
an executable package generated from one of a plurality of machine learning models;
a non-machine learning model-based executable package; or
an executable package generated from a composite solution comprising at least one other artifact.

17. A non-transitory computer-readable storage medium storing instructions which, when executed by a processing system including at least one processor, cause the processing system to perform operations, the operations comprising:
receiving a first machine learning model, wherein the first machine learning model is in a first format associated with a first development environment;
adapting the first machine learning model to a containerized environment, wherein the adapting the first machine learning model to the containerized environment comprises:
generating a first artifact associated with the first machine learning model, wherein the first artifact includes library information for deployment of the first machine learning model in the containerized environment; and
defining at least a second artifact associated with the first machine learning model, wherein the at least the second artifact defines an external compatibility of at least one of an input port or an output port of the first machine learning model;

validating the first machine learning model according to at least one validation criterion associated with a repository; and
publishing the first machine learning model to the repository.

18. A device comprising:
a processing system including at least one processor; and
a computer-readable medium storing instructions which, when executed by the processing system, cause the processing system to perform operations, the operations comprising:
receiving a first machine learning model, wherein the first machine learning model is in a first format associated with a first development environment;
adapting the first machine learning model to a containerized environment, wherein the adapting the first machine learning model to the containerized environment comprises:
generating a first artifact associated with the first machine learning model, wherein the first artifact includes library information for deployment of the first machine learning model in the containerized environment; and
defining at least a second artifact associated with the first machine learning model, wherein the at least the second artifact defines an external compatibility of at least one of an input port or an output port of the first machine learning model;

validating the first machine learning model according to at least one validation criterion associated with a repository; and
publishing the first machine learning model to the repository.

19. The device of claim 18 , wherein the validating the first machine learning model according to the at least one validation criterion associated with the repository comprises:
validating the first artifact and the second artifact associated with the first machine learning model.

20. The device of claim 18 , wherein the operations further comprise:
validating a composite solution including the first machine learning model according to the at least one validation criterion associated with the repository.

Assignees

• At & T Ip I Lp

Inventors

• Aftab Syed Anwar
• Jacobson Guy
• Klein Reuben
• Murray John
• Gilbert Mazin
• Talasila Manoop
• Farooqui Kazi

Classifications

• H04L67/10

  in which an application is distributed across nodes in the network (software deployment G06F8/60; multiprogramming arrangements G06F9/46) · CPC title

• G06N20/00 Primary

  Machine learning · CPC title

• H04L67/02

  based on web technology, e.g. hypertext transfer protocol [HTTP] · CPC title

• G06F9/45533

  Hypervisors; Virtual machine monitors · CPC title

• G06N5/02

  Knowledge representation; Symbolic representation · CPC title