Remote ML Arrangement for Shared Training

Solution Overview

Problem

Enterprises face challenges in evaluating data stored in cloud-based networks due to the high computational resource consumption and cost associated with machine learning (ML) software, which can be costly and resource-intensive.

Innovation Solution

A cloud-based network system that provides a remote ML arrangement allowing multiple enterprises to share ML models and predictions securely, using a computing system and trainer devices to generate ML models based on enterprise data, reducing the need for dedicated computational resources and costly software.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If enterprises use dedicated ML software to evaluate cloud-based data, then data evaluation capability is improved, but computational resource consumption and cost increase

Engineering Contradiction:

Improvedata evaluation capabilityVSAvoidcomputational resource consumption

Core Design Contradiction:

ReliabilityVSUse of energy by moving object

Solution Approach 1:

Multiple enterprises share a common remote ML arrangement and trainer device pool, merging their individual ML needs into a unified infrastructure. This allows enterprises to access ML capabilities without maintaining separate dedicated systems, reducing overall computational resource consumption while maintaining data evaluation capability.

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The remote ML arrangement is designed to serve multiple enterprises simultaneously, making the trainer devices and ML infrastructure universal rather than dedicated to a single enterprise. This multi-functional system can handle different ML training requests from various enterprises, reducing the need for each enterprise to maintain separate computational resources.

Inventive Principle:

Principle #6Universality (Multi-functionality)

2Reliability

If enterprises use dedicated ML software to evaluate cloud-based data, then data evaluation capability is improved, but cost increases

Engineering Contradiction:

Improvedata evaluation capabilityVSAvoidcost

Core Design Contradiction:

ReliabilityVSQuantity of substance

Solution Approach 1:

Multiple enterprises share a common remote ML arrangement and trainer device pool, merging their individual ML needs into a unified infrastructure. This allows enterprises to access ML capabilities without maintaining separate dedicated systems, reducing overall computational resource consumption while maintaining data evaluation capability.

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The remote ML arrangement is designed to serve multiple enterprises simultaneously, making the trainer devices and ML infrastructure universal rather than dedicated to a single enterprise. This multi-functional system can handle different ML training requests from various enterprises, reducing the need for each enterprise to maintain separate computational resources.

Inventive Principle:

Principle #6Universality (Multi-functionality)

3Productivity

If enterprises deploy ML trainer processes locally, then ML model generation speed is improved, but device complexity and resource requirements increase

Engineering Contradiction:

ImproveML model generation speedVSAvoidsystem complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

A remote ML arrangement acts as an intermediary between enterprises and the actual ML training processes. Instead of enterprises directly deploying and managing complex trainer processes locally, they submit training requests to the remote system, which handles the complex ML model generation using shared trainer devices. This maintains productivity while reducing the complexity burden on individual enterprises.

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS11620571B2Machine learning with distributed training

Publication Date: 2023.04.04 SERVICENOW INC

AI summary

A network system may include a plurality of trainer devices and a computing system disposed within a remote network management platform. The computing system may be configured to: receive, from a client device of a managed network, information indicating (i) training data that is to be used as basis for generating a machine learning (ML) model and (ii) a target variable to be predicted using the ML model; transmit an ML training request for reception by one of the plurality of trainer devices; provide the training data to a particular trainer device executing a particular ML trainer process that is serving the ML training request; receive, from the particular trainer device, the ML model that is generated based on the provided training data and according to the particular ML trainer process; predict the target variable using the ML model; and transmit, to the client device, information indicating the target variable.