Information Processing Device Optimizing Solution Diversity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing optimization systems struggle to output diverse solutions effectively, often resulting in multiple solutions with close distances when aiming for diversity in optimization problems.
Innovation Solution
An information processing device and method that select and output a predetermined number of solutions where the minimum distance between them satisfies a predetermined condition, using a Multiplicative Weight Update algorithm to maximize the minimum weighted Hamming distance, preventing similarity among output solutions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If diverse solutions are sought by maximizing sum of distances, then the number of diverse solutions increases, but solutions with close distance may still be included
Solution Approach 1:
The patent changes the optimization parameter from maximizing the sum of distances to maximizing the minimum distance between solutions. This parameter transformation ensures that all output solutions maintain a guaranteed minimum separation, preventing similar solutions from being selected together while preserving solution diversity.
Solution Approach 2:
The patent applies a preprocessing step that computes distance matrices and identifies candidate solutions satisfying the minimum distance condition before the main selection process. This beforehand cushioning ensures that the diversity constraint is built into the solution set from the beginning, preventing the inclusion of similar solutions.
2Productivity
If a predetermined number of solutions are output, then the output quantity is controlled, but diversity among solutions cannot be guaranteed
Solution Approach 1:
The patent performs preliminary actions by pre-computing distance matrices and identifying candidate solution sets that satisfy the minimum distance condition before the actual solution selection. This preliminary processing ensures that when solutions are output, diversity is already guaranteed by the candidate selection criteria.
Solution Approach 2:
The patent uses feedback mechanisms where the distance between solutions is continuously evaluated during the selection process. The minimum distance constraint acts as a feedback criterion that guides the selection of subsequent solutions, ensuring that each new solution maintains the required diversity from previously selected solutions.
Data Source
AI summary
An information processing device 1X mainly includes an output means 15X. The output means outputs, as a predetermined number of solutions, candidates in which a minimum value of a distance between the solutions satisfies a predetermined condition among candidates used to select the predetermined number of solutions, when outputting the predetermined number of solutions, from a plurality of solutions in an optimization problem.


