Neural Network Genotype Phenotype Segmentation for Search Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems using genetic algorithms in neural networks lack discrimination between phenotype and genotype, leading to low search efficiency and increased system size, making them costly and impractical for small companies or development sites.
Innovation Solution
A cell differentiation algorithm is applied to a neural network to recognize environmental conditions, allowing for initial settings to be evaluated and optimized, thereby improving search efficiency and reducing unnecessary processes, enabling the system to operate effectively in smaller environments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a genetic algorithm is applied to a neural network without discrimination between phenotype and genotype, then the system can derive optimal techniques through repeated learning, but search efficiency is low and the number of nodes increases resulting in increased search time
Solution Approach 1:
The patent segments the search process into two distinct phases: genotype search (optimizing network structure and parameters) and phenotype search (optimizing operational behaviors). This segmentation allows each phase to focus on specific optimization tasks, improving overall search efficiency by avoiding redundant searches and reducing the total number of nodes required in the neural network.
Solution Approach 2:
The patent extracts and separates the genotype and phenotype components from the unified neural network structure. By taking out the genotype (network architecture and parameters) and phenotype (operational behaviors) as distinct elements, the system can optimize them independently through targeted search processes, thereby reducing search time and improving efficiency.
2Measurement precision
If a considerable number of searches for a long time are performed, then accurate results can be derived, but the system size increases making it costly and impractical for small companies or development sites
Solution Approach 1:
The patent applies preliminary action by pre-defining the neural network structure and parameters (genotype) before the actual operational search (phenotype). This preliminary setup allows the system to start with a optimized framework, reducing the computational burden during runtime and enabling accurate results with a smaller, more cost-effective system configuration suitable for small companies and development sites.
Data Source
AI summary
Disclosed are an optimal technique search method and system that can enable a more effective search for optimal techniques for problem solutions than in the past through the use of a neural network employing genetic algorithm. Provided therein are an execution unit (1) that uses a neural network employing a genetic algorithm to search for an optimal technique and which executes operations using said technique, and an evaluation unit (2) that, along with creating initial setting to transmit to said execution unit, evaluates the content of the operations of the execution unit after the operations have been executed and has the execution unit (1) execute operations a plurality of times, and thereby derives as the optimal technique the initial settings that executed the most effective operation when transmitted to the execution unit (1) out of the results derived from said plurality of operation executions. As a result, a small scale and effective optimal technique search becomes possible when using a neural network, as described in [0024] and [0025].