Quantized Local Field Updates for Ising Inequality Constraints
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for solving discrete optimization problems with inequality constraints using Ising apparatuses face inefficiencies in arithmetic operations due to the complexity of handling discontinuous linear forms.
Innovation Solution
A data processing apparatus that utilizes a storage unit to store total energy and local fields, and a processing unit to update these fields using quantized local fields to efficiently handle inequality constraints, thereby reducing arithmetic operation costs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a discontinuous linear form constraint term is used to represent inequality constraints in an Ising model, then the constraint condition can be expressed in the evaluation function, but the arithmetic operation efficiency deteriorates due to complexity in handling the discontinuous linear form
Solution Approach 1:
The patent changes the parameter representation by introducing a quantized local field parameter that approximates the discontinuous linear form constraint term. This quantized parameter transforms the complex discontinuous function into a manageable discrete representation, enabling efficient arithmetic operations while preserving the constraint representation capability.
Solution Approach 2:
The patent uses a simplified quantized local field parameter that may lose some precision but significantly reduces computational complexity. This approximate parameter serves the purpose of constraint representation efficiently, sacrificing minor precision for major gains in arithmetic operation speed.
2Measurement precision
If the local field is updated using the full precision second local field, then the solution accuracy is maintained, but the arithmetic operation complexity and processing time increase
Solution Approach 1:
The patent changes the parameter used in local field updates from the full precision second local field to a quantized version. This parameter transformation maintains sufficient solution accuracy while dramatically reducing the arithmetic operation complexity and processing time required for updates.
Solution Approach 2:
The patent applies partial action by using quantized approximation instead of full precision calculations. This partial approach provides sufficient accuracy for the optimization problem while avoiding the excessive computational cost of using the complete second local field values in all calculations.
3Productivity
If quantized local fields are used to update the first local field, then the arithmetic operation efficiency is improved, but the solution precision may be affected
Solution Approach 1:
The patent changes the parameter representation to a quantized form that balances precision and efficiency. The quantized local field parameter maintains enough information to achieve accurate solutions while enabling efficient arithmetic operations, thus resolving the contradiction between productivity and measurement precision.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A data processing apparatus configured to: acquire total energy of a constraint term, values of a plurality of state variables included in an Ising-type evaluation function, a first weight value between the plurality of state variables, a second weight value between state variables and the constraint condition, a first local field of the total energy of changed each of the plurality of state variables, and a second local field used for determination of a constraint violation amount, repeat determining whether a change of a first state variable is allowed, repeat, when allowed, updating the first local field based on the first weight value and a first quantized local field, updating the second local field based on the second weight and a second quantized local field, and search for a combination of values of the plurality of state variables whose value is local value.