Integer-Array Optimization with One-Way One-Hot Constraints
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing binary variable sampling techniques for optimization, such as those using Ising models in the QUBO format, suffer from inefficient sampling and increased sampling times to achieve high accuracy solutions, and methods like FMQA and FMDA face limitations in expressing complex problems and adjusting recommendations.
Innovation Solution
An arithmetic program and information processing device that generates initial points in an integer array format, applies a one-way one-hot constraint, and uses a combination of FMDA and GA to optimize the sampling process, ensuring the one-way one-hot constraint is satisfied, thereby reducing sampling times while maintaining high accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If binary variable sampling techniques using Ising models in QUBO format are used for optimization, then optimization capability is provided, but sampling efficiency is low and sampling times increase to achieve high accuracy solutions
Solution Approach 1:
The patent segments the sampling process into two distinct phases: (1) generating initial points that satisfy the one-way one-hot constraint, and (2) performing optimization sampling from these constrained initial points. This segmentation allows the system to first ensure constraint satisfaction, then focus sampling efforts on finding high-accuracy solutions, thereby reducing overall sampling time while maintaining accuracy.
Solution Approach 2:
The patent applies preliminary action by generating initial points that already satisfy the one-way one-hot constraint before performing the actual optimization sampling. By pre-establishing constraint-compliant starting points, the system avoids wasting sampling iterations on invalid solutions, thus reducing the number of sampling times needed to achieve high accuracy.
2Adaptability or versatility
If methods like FMQA and FMDA are used for optimization, then quantum annealing or digital annealing is applied, but they face limitations in expressing complex problems and adjusting recommendations
Solution Approach 1:
The patent introduces an intermediary component: the one-way one-hot constraint mechanism that bridges the training data and the optimization process. This intermediary structure enables complex problem expression by systematically organizing training data into constraint-compliant initial points, allowing FMQA and FMDA methods to handle more complex problems without increasing their inherent computational complexity.
3Productivity
If traditional sampling methods are used without constraint satisfaction, then sampling process is simpler, but local solutions are obtained and sampling efficiency decreases
Solution Approach 1:
The patent applies preliminary anti-action by proactively preventing the system from entering local solution traps through the one-way one-hot constraint. By designing initial points that inherently satisfy constraints, the system counteracts the tendency to converge to suboptimal local solutions before the sampling process begins, thereby improving both sampling efficiency and solution reliability.
Data Source
AI summary
A computer-readable recording medium stores a program for causing a computer to execute a process including: generating an integer array of types as an initial point of each training data of the training data group; searching for the first point by providing a constraint in which, when the first point is retrieved, an index of a variable in the integer array is set to i and an index that represents a type of the variable is set to j for each training data, a case of i=j in the integer array in a matrix of i and j is set to 1 and another case is set to 0 to perform conversion into a bit array of i and j, and variable in each row is 1 in the matrix; and applying the genetic algorithm to a format of the integer array when the second point is retrieved.