Correlation-Aware Binary Allocation for Ising Model Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Converting combinatorial optimization problems with continuous variables into Ising models using binary expansion results in a large number of binary variables, leading to increased problem scale and reduced accuracy when converting back to continuous variables.
Innovation Solution
A data processing device that detects the correlation between continuous variables and allocates a larger number of common binary variables to correlated pairs, generating correspondence information to convert the evaluation function into an Ising-type evaluation function with reduced bit usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If binary expansion is used to convert continuous variables into binary variables, then the combinatorial optimization problem can be solved using Ising model, but the number of binary variables significantly increases leading to larger problem scale
Solution Approach 1:
The patent merges continuous variables with high correlation into common binary variable groups. Instead of allocating separate binary variables to each continuous variable, the system identifies correlated variables and shares binary variables among them, thereby reducing the total number of binary variables while preserving the ability to represent the solution space accurately.
Solution Approach 2:
The patent applies different bit allocation strategies to different continuous variables based on their local characteristics, specifically their correlation with other variables. Variables with high correlation receive fewer bits through shared allocation, while independent variables receive full bit allocation, optimizing the overall problem scale while maintaining solution accuracy.
2Device complexity
If the number of bits allocated to each continuous variable is reduced, then the problem scale decreases, but the accuracy of the obtained solution is lowered
Solution Approach 1:
By merging continuous variables that share high correlation into common binary variable groups, the patent reduces the total number of binary variables required. This merging allows the system to maintain adequate precision for representing continuous variables while significantly reducing the overall problem scale, as the same binary variables serve multiple correlated continuous variables simultaneously.
3Ease of manufacture
If simple binary expansion is used for conversion, then the conversion process is simple, but the number of bits significantly increases depending on the number of continuous variables
Solution Approach 1:
The patent performs preliminary analysis of correlation among continuous variables before the binary conversion process. By pre-identifying which variables are highly correlated, the system can proactively plan shared binary variable allocation, avoiding the need for simple independent binary expansion and thereby reducing the total number of binary variables from the outset.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A data processing device configured to: detect magnitude of correlate on of a continuous variable pair included in a plurality of continuous variables based on information regarding a first evaluation function that includes the plurality of continuous variables obtained by formulating a combinatorial optimization problem, allocate a larger number of common binary variables to the continuous variable pair as the correlation is larger at a time of allocating a binary variable to each of the plurality of continuous variables, generate correspondence information that indicates a correspondence relationship between each of the plurality of continuous variables and the binary variable, convert the first evaluation function into a second evaluation function that includes a plurality of binary variables, the second evaluation function being Ising-type, set coefficient information of the second evaluation function, and search for a solution to the combinatorial optimization problem using the second evaluation function and the coefficient information.