Ising Model Segmentation for Multi-Characteristic Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In optimization tasks involving a plurality of characteristic values, existing models may fail to accurately reflect the features of each characteristic value, leading to decreased model accuracy and increased sampling requirements.
Innovation Solution
An arithmetic program that repeatedly creates a plurality of Ising models for each characteristic value based on a training data group, combines these models into a single Ising model, and iteratively updates the training data with recommended points from the combined model to improve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single Ising model is created to handle multiple characteristic values, then the model structure remains simple, but the model accuracy decreases because features of each characteristic value are not properly reflected
Solution Approach 1:
The patent segments the modeling process by creating separate Ising models for each characteristic value instead of using a single aggregated model. This segmentation allows each characteristic value's features to be captured independently in its own model, thereby improving model accuracy while maintaining manageable complexity through systematic combination of individual models.
2Speed
If sampling is performed with lower model accuracy, then the sampling process is faster, but more sampling iterations are required to achieve optimal solutions
Solution Approach 1:
The patent implements feedback by iteratively updating the training data group with newly acquired data and characteristic values from each sampling iteration. This feedback mechanism allows the model to learn from previous sampling results, improving model accuracy over time and reducing the total number of iterations needed, thereby enhancing overall sampling efficiency.
Data Source
Figure 1
Figure 2
Figure 3A~3B
AI summary
An arithmetic program including instructions which, when executed by a computer, cause the computer to repeatedly execute processing including: creating a plurality of Ising models by creating the Ising models for each of a plurality of characteristic values, based on a training data group that includes a plurality of pieces of training data in which data and the plurality of characteristic values for the data are associated with each other; creating a combined Ising model by combining the plurality of Ising models; searching for the data of a recommended point for the combined Ising model; and adding, to the training data group, the training data that includes the data of the recommended point and the plurality of characteristic values for the data of the recommended point.