Graphical Machine Learning Platform Component Reuse Mechanism
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users of graphic machine learning algorithm platforms need to re-establish functional models repeatedly to access previously used functions, which is inefficient.
Innovation Solution
A method to release an established functional model as a new component, allowing users to directly select and reuse the new component without re-establishing the model, by determining input and output ends and unique identifiers of mandatory parameters, and validating the component's functionality through test data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a functional model is re-established repeatedly to access previously used functions, then the model can be rebuilt with current data, but the time consumption and operational efficiency deteriorate
Solution Approach 1:
The patent creates a copy of the functional model as a reusable component. Instead of re-establishing the model from scratch, the system copies the existing model structure, parameters, and functionality into a new component that can be directly selected and reused in different modeling scenarios, significantly reducing time consumption while maintaining model reliability
Solution Approach 2:
The patent segments the functional model into a standalone component with defined input and output interfaces. This segmentation allows the model to be extracted as an independent unit that can be reused across different contexts without requiring re-configuration, thereby reducing the time needed for repeated model establishment
2Reliability
If a functional model is re-established repeatedly, then the model can be optimized with new data, but the ease of operation deteriorates
Solution Approach 1:
The system creates a reusable component copy that preserves the model's optimized parameters and structure. Users can directly select and instantiate this pre-optimized component without needing to re-establish the model, significantly improving ease of operation while maintaining the model's accuracy through the copied parameters
Solution Approach 2:
The patent performs preliminary actions by pre-optimizing the model and packaging it as a component with predefined interfaces and parameters. This preliminary preparation eliminates the need for users to perform time-consuming model re-establishment operations, thereby improving ease of operation while ensuring the model maintains its optimized accuracy
3Measurement precision
If components are created with unique identifiers for mandatory parameters, then parameter identification accuracy improves, but the device complexity increases
Solution Approach 1:
The patent extracts the parameter identification function into a separate mechanism using unique identifiers. By taking out the complex parameter matching logic and replacing it with simple unique identifier matching, the system achieves high parameter identification accuracy without significantly increasing overall device complexity
Solution Approach 2:
The system copies the parameter identification mechanism across all components using standardized unique identifiers. This copying approach allows consistent and accurate parameter identification throughout the system without requiring complex custom identification logic for each component, thereby maintaining low complexity while achieving high precision
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A component release method, a graphic machine learning algorithm platform-based component building method, and a graphic machine learning algorithm platform are provided. The graphic machine learning algorithm platform can release or build a functional model as a new component. Therefore, when a user needs to use a certain function repeatedly, the new component can be directly used without the need of repeatedly building a functional model, thus facilitating use by the user.