Computing Experiment System Pruning Non-Relevant Jobs

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing methods for evaluating 'what if' scenarios and calibrating simulation models face challenges in determining which parameters to evaluate first and how to change them efficiently to maximize cost- and time-efficiency, particularly in non-trivial activities like scheduling, routing, and optimization problems.

Innovation Solution

A system and method for conducting computing experiments that execute a set of jobs, compare results with templates of previously-executed experiments stored in a knowledge base, and identify prunable jobs based on user constraints, allowing for guided experimentation that maximizes business impact by modeling user behavior and providing feedback.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Extent of automation

If fully automatic solutions are adopted to solve non-trivial activities such as scheduling, routing, and optimization problems, then the extent of automation is improved, but the time and cost efficiency deteriorates due to the need to evaluate numerous parameters and escape from non-optimal outputs

Engineering Contradiction:

ImproveautomationVSAvoidtime efficiency

Core Design Contradiction:

Extent of automationVSLoss of time

Solution Approach 1:

The system performs preliminary actions by executing a subset of jobs first to generate result data before committing to full experiment execution. This preliminary evaluation allows the system to identify promising parameter combinations early, avoiding wasted time on non-optimal paths while maintaining automation.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system implements feedback mechanisms by continuously comparing result data from executed jobs against thresholds and user constraints. This feedback loop enables dynamic adjustment of the experiment execution plan, pruning non-promising branches and redirecting resources to more promising parameter combinations, thereby improving time efficiency without reducing automation.

Inventive Principle:

Principle #23Feedback

2Extent of automation

If fully automatic solutions are adopted to solve non-trivial activities such as scheduling, routing, and optimization problems, then the extent of automation is improved, but the cost efficiency deteriorates due to the need to evaluate numerous parameters and escape from non-optimal outputs

Engineering Contradiction:

ImproveautomationVSAvoidcost efficiency

Core Design Contradiction:

Extent of automationVSLoss of energy

Solution Approach 1:

The system performs preliminary actions by executing a subset of jobs first to generate result data before committing to full experiment execution. This preliminary evaluation allows the system to identify promising parameter combinations early, avoiding wasted computational resources on non-optimal paths while maintaining automation.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

Inventive Principle:

Principle #23Feedback

3Measurement precision

If all jobs in a set are executed without pruning, then the measurement precision is improved by evaluating all parameter combinations, but the loss of time increases due to the inability to skip non-relevant jobs

Engineering Contradiction:

Improveevaluation completenessVSAvoidexperiment duration

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system executes a preliminary subset of jobs to generate result data that informs subsequent execution decisions. This preliminary action maintains measurement precision by ensuring that promising parameter combinations are thoroughly evaluated while avoiding time waste on non-promising combinations.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system applies partial action by executing only the necessary subset of jobs rather than all possible parameter combinations. By using feedback from preliminary executions to prune non-promising branches, the system achieves sufficient measurement precision for decision-making without the excessive time cost of exhaustive evaluation.

Inventive Principle:

Principle #16Partial or excessive action

4Measurement precision

If all jobs in a set are executed without pruning, then the measurement precision is improved by evaluating all parameter combinations, but the loss of computational resources increases

Engineering Contradiction:

Improveevaluation completenessVSAvoidcomputational resource usage

Core Design Contradiction:

Measurement precisionVSUse of energy by moving object

Solution Approach 1:

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system implements feedback mechanisms by continuously comparing result data from executed jobs against thresholds and user constraints. This feedback loop enables dynamic adjustment of the experiment execution plan, pruning non-promising branches and redirecting computational resources to more promising parameter combinations, thereby improving resource efficiency while maintaining evaluation quality.

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS12032877B2System and method for conducting computing experiments

Publication Date: 2024.07.09 INTERNATIONAL BUSINESS MACHINE CORPORATION

AI summary

A method of conducting computing experiments includes executing a set of jobs based on user-selected parameters, learning a user strategy by checking the user-selected parameters during the executing of the set of jobs, and refining the user strategy by refining the set of jobs.