Providing custom machine-learning models

Patent metadata

Field	Value
Publication number	US-12154015-B2
Application number	US-202217963028-A
Country	US
Kind code	B2
Filing date	Oct 10, 2022
Priority date	Nov 29, 2018
Publication date	Nov 26, 2024
Grant date	Nov 26, 2024

Abstract

Providing custom machine learning models to client computer systems. Multiple machine learning models are accessed. Client-specific data for multiple client computer systems are also accessed. For each of at least some of the client computer systems, performing the following actions: First, using the corresponding client-specific data for the corresponding client computer system to determine which subset of the multiple machine learning models is applicable to the corresponding client computer system. The subset of the multiple machine learning models includes more than one of the multiple machine learning models. Then, aggregating the determined subset of the multiple machine learning models to generate an aggregated subset of machine learning models that is customized to the corresponding client computer system.

First claim

What is claimed is:

1. A method for providing custom machine learning models to a computer system, said method comprising:
accessing data for the computer system;
using the data to select a set of machine learning models that are determined to be applicable to the computer system, wherein the computer system is facilitating development of a codebase in a programming language such that machine learning models that are directed to codebase development are ones that are determined to be applicable to the computer system, and wherein determining which machine learning models are to be included in the set is performed by:
determining attributes associated with the data;
accessing a plurality of machine learning models;
using the attributes to filter the plurality of machine learning models, resulting in identification of machine learning models that are directed to codebase development, wherein said filtering involves identification of a particular machine learning model that detects variable misuses for codebases that are developed in said programming language;
using machine learning to build an additional model that is applicable to the computer system based on the determined attributes; and
including at least the built model and the particular machine learning model in the set of machine learning models; and

aggregating the set to generate an aggregated set of machine learning models that are customized to the computer system.

2. The method of claim 1 , wherein said filtering is further based on a type of an application being designed or implemented by the computer system.

3. The method of claim 1 , wherein the aggregate set of machine learning models is aggregated into a single integrated model that is customized to the computer system.

4. The method of claim 1 , wherein the computer system operates in a programming environment that includes language specific editors and debugging facilities.

5. The method of claim 1 , wherein the computer system operates in multiple programming environments, and wherein separate data is accessible for each of the multiple programming environments.

6. The method of claim 1 , wherein the aggregate set of machine learning models is aggregated into a single integrated model that is customized to the computer system, and
wherein the single integrated model is caused to operate in a split manner so that only a part of the single integrated model operates for the computer system.

7. The method of claim 1 , wherein a first model included in the set of machine learning models is a base model.

8. The method of claim 7 , wherein a second model included in the set of machine learning models is a custom model.

9. The method of claim 8 , wherein one of the base model or the custom model is generated from a public codebase.

10. One or more hardware storage devices that store instructions that are executable by one or more processors to cause the one or more processors to:
access a plurality of machine learning models;
access data for a computer system;
use the data to determine which subset of the plurality of machine learning models is applicable to the computer system, wherein the computer system is facilitating development of a codebase in a programming language such that machine learning models that are directed to codebase development are ones that are determined to be applicable to the computer system, and wherein determining which machine learning models are to be included in the subset is performed by:
determining attributes associated with the data;
using the attributes to filter the plurality of machine learning models, resulting in identification of machine learning models that are directed to codebase development, wherein said filtering involves identification of a particular machine learning model that detects variable misuses for codebases that are developed in said programming language;
using machine learning to build an additional model that is applicable to the computer system based on the determined attributes; and
including at least the built model and the particular machine learning model in the subset; and

aggregate the subset to generate an aggregated set of machine learning models that are customized to the computer system.

11. The one or more hardware storage devices of claim 10 , wherein a first model is included in the subset, and wherein the first model is generated for a particular team.

12. The one or more hardware storage devices of claim 10 , wherein a first model is included in the subset, and wherein the first model is generated for a software developer kit (SDK).

13. The one or more hardware storage devices of claim 10 , wherein a first model is included in the subset, and wherein the first model is subject to a codebase owner's control.

14. The one or more hardware storage devices of claim 10 , wherein determining which machine learning models are to be included in the subset is based on a permissions factor.

15. The one or more hardware storage devices of claim 14 , wherein the permission factor is set by a client or a service provider.

16. The one or more hardware storage devices of claim 14 , wherein a service provider sets a scope of permissions based on a determined subscription service associated with the computer system.

17. The one or more hardware storage devices of claim 10 , wherein determining which machine learning models are to be included in the subset is based on a group membership determination.

18. A computer system that provides custom machine learning models to a computer system, said computer system comprising:
at least one processor; and
at least one hardware storage device having stored thereon instructions that are executable by the at least one processor to cause the computer system to:
access data for a device;
use the data to select a set of machine learning models that are determined to be applicable to the device, wherein the computer system is facilitating development of a codebase in a programming language such that machine learning models that are directed to codebase development are ones that are determined to be applicable to the computer system, and wherein determining which machine learning models are to be included in the set is performed by:
determining attributes associated with the data;
accessing a plurality of machine learning models;
using the attributes to filter the plurality of machine learning models, resulting in identification of machine learning models that are directed to codebase development, wherein said filtering involves identification of a particular machine learning model that detects variable misuses for codebases that are developed in said programming language;
using machine learning to build an additional model that is applicable to the device based on the determined attributes; and
including at least the built model and the particular machine learning model in the set of machine learning models; and

aggregate the set to generate an aggregated set of machine learning models that are customized to the device.

19. The computer system of claim 18 , wherein a first model in the set of machine learning models is designed to write programming code.

20. The computer system of claim 18 , wherein a first model included in the set of machine learning models is a base model or a custom model.

Assignees

• Microsoft Technology Licensing Llc

Inventors

• Keech Jonathan Daniel
• Shanmugam Kesavan
• Calvert Simon
• Wilson-Thomas Mark A
• Lim Vivian Julia

Classifications

• G06F8/33

  Intelligent editors · CPC title

• G06F8/436

  Semantic checking · CPC title

• G06N20/20 Primary

  Ensemble learning · CPC title