GNN Prediction Model for Mixed Real and Simulation Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional technologies using Graph Neural Networks (GNN) are limited in their ability to integrate and process various types of data sets such as actual data, simulation result data, experimental result data, and calculation result data, preventing the acquisition of accurate prediction results across different data types.
Innovation Solution
A prediction apparatus that utilizes a learning model, including a graph neural network (GNN), to process and integrate multiple types of data sets, enabling the acquisition of prediction results by performing machine learning prediction processing on actual data, simulation results, and experimental or calculation results, with the capability to predict future data based on historical data sets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional GNN technology is used, then graph structure processing capability is provided, but the ability to integrate and process multiple types of data sets (actual data, simulation result data, experimental result data, and calculation result data) is limited
Solution Approach 1:
The patent creates a unified learning model that can process multiple types of data sets (actual data, simulation result data, experimental result data, and calculation result data) through a single GNN framework. The model accepts diverse data types as input and generates predictions for any of these data types, making the system multi-functional and adaptable to different data sources while maintaining reliable prediction accuracy across all data types.
2Ease of manufacture
If a single data set type is used for training, then the learning model is simple to create, but it is impossible to acquire other types of data sets
Solution Approach 1:
The patent merges multiple data set types into a single unified learning model. Instead of creating separate models for each data type, the invention combines the processing capabilities for actual data, simulation result data, experimental result data, and calculation result data into one GNN model that can handle all these data types simultaneously, achieving both simplicity and versatility.
3Adaptability or versatility
If multiple types of data sets are integrated into one learning model, then diverse data processing capability is achieved, but the device complexity increases
Solution Approach 1:
The patent employs a universal GNN learning model that handles multiple data types through a single unified architecture. This approach avoids the complexity of creating and maintaining multiple separate models, as the single multi-functional model can process actual data, simulation result data, experimental result data, and calculation result data while generating predictions for any of these types.
Data Source
AI summary
A prediction apparatus includes: an acceptance unit that accepts a group of explanatory variables constituted by one or more types of data sets selected from the group consisting of four types of data sets, namely, an actual data set, a simulation result set, an experimental result set, and a calculation result set for a given subject; a prediction unit that acquires a prediction result by performing machine learning prediction processing, using a learning model created by performing learning processing using a group of training data including two or more types of data sets selected from the group consisting of an actual data set, a simulation result set, an experimental result set, and a calculation result set for the given subject, and the accepted group of explanatory variables; and an output unit that outputs the prediction result acquired by the prediction unit.