ML Workflow Branch Selection for Target Indicator Planning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing measure planning in government health services often relies on intuition and assumptions, making it difficult to generate workflows aimed at achieving target indicators, and existing prediction technologies are slow and inefficient in calculating the effects of measure modifications.
Innovation Solution
A computer-implemented workflow generation method using a machine learning model to calculate indicator values for candidate items, select items that satisfy predetermined conditions, and generate efficient workflows by referring to a storage of candidate items and their conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If measure planning is based on intuition and assumptions, then the planning process is simple and quick, but the accuracy of achieving target indicators is low
Solution Approach 1:
The patent introduces a prediction model as an intermediary tool between the planner's intentions and the actual measure implementation. This model calculates indicator values for different candidate items, providing data-driven guidance that improves planning accuracy without requiring complete redesign of the planning process
Solution Approach 2:
The system performs preliminary calculations of indicator values for multiple candidate items before final decision-making. By pre-evaluating the effects of different measures using the prediction model, the system enables more informed decisions while maintaining a structured yet flexible planning workflow
2Productivity
If existing prediction technologies are used to calculate effects of measure modifications, then prediction accuracy is achieved, but calculation speed is slow and efficiency is low
Solution Approach 1:
The patent segments the prediction process into distinct components: acquiring candidate items, calculating indicator values for each candidate, and selecting optimal items. This segmentation allows for efficient processing by focusing calculations only on relevant candidates rather than exhaustive analysis
Solution Approach 2:
The system dynamically adjusts calculation parameters based on the specific planning context, such as which candidate items are being evaluated and what indicator types are needed. This parameter optimization enables faster calculations by avoiding unnecessary computations while maintaining prediction accuracy
Data Source
AI summary
A computer-implemented workflow generation method in which a computer executes processing of calculating, in a case where a conditional branch item included in a workflow is acquired, an indicator value in a case where any one of a plurality of candidate items is used by using a machine learning model, by referring to a storage that stores the plurality of candidate items satisfying a condition of a branch destination corresponding to the conditional branch item and the machine learning model that calculates the indicator value of each of the plurality of candidate items, selecting an item of which the calculated indicator value satisfies a predetermined condition from among the plurality of candidate items, and generating a workflow in which the selected item is arranged at the branch destination of the conditional branch item.