Generic Evolution Service for Multi-Domain Data Structures

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing methods for evolutionary computation lack a generic system capable of evolving in a secure and generic manner across disparate optimization domains, requiring specialized code and failing to adapt evolutionary approaches to different problems effectively.

Innovation Solution

A system that allows specifying data descriptions for evolving data structures in a unified form, enabling secure evolution of models across various domains by using configuration specifications to generate genomes, and translating them into a representation suitable for fitness evaluation, while maintaining security and intellectual property protection.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If specialized code is written for performing evolution for specific problem domains, then the evolution can be performed effectively for that domain, but the system lacks the ability to retarget evolutionary approaches to different problems and requires reimplementation for each domain

Engineering Contradiction:

Improveeffectiveness of evolution for specific domainVSAvoidability to retarget evolutionary approaches to different problems

Core Design Contradiction:

ReliabilityVSAdaptability or versatility

Solution Approach 1:

The patent implements a universal evolution service that can handle multiple disparate optimization domains through a single generic system. The system uses a unified configuration specification format that can describe various data structures (neural networks, decision trees, etc.) and evolutionary operators, allowing the same core evolution engine to serve multiple domains without specialized code for each problem type.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The system changes the approach from domain-specific code to domain-agnostic configuration. By parameterizing the evolution process through configuration specifications that define data structures, operators, and fitness functions, the system can adapt to different domains by changing parameters rather than rewriting code, thus achieving both reliability and versatility.

Inventive Principle:

Principle #35Parameter changes

2Adaptability or versatility

If a generic evolution system is created to serve multiple domains, then adaptability improves, but the system becomes complex to handle disparate genetic material models and data sets

Engineering Contradiction:

Improveability to serve multiple optimization domainsVSAvoidcomplexity of handling disparate genetic material models

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent segments the evolution system into distinct modular components: configuration specification parsing, population management, fitness evaluation, and result interpretation. Each component handles a specific aspect of the evolution process independently, working with domain-specific data through standardized interfaces. This segmentation reduces complexity by localizing domain-specific handling to configuration files rather than requiring complex integration logic throughout the system.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The configuration specification acts as an intermediary layer between the generic evolution engine and domain-specific problems. It translates domain-specific requirements into a unified format that the evolution service can process, eliminating the need for the system to directly handle disparate genetic material models and data sets, thus reducing complexity while maintaining versatility.

Inventive Principle:

Principle #24Intermediary (Mediator)

3Ease of operation

If configuration specifications are used to define evolution parameters, then the ability to change dimensionality and representational complexity as configuration improves, but the system requires robust parsing and validation mechanisms

Engineering Contradiction:

Improveability to change dimensionality and complexity as configurationVSAvoidparsing and validation mechanisms required

Core Design Contradiction:

Ease of operationVSDevice complexity

Solution Approach 1:

The system performs preliminary validation and parsing of configuration specifications before the evolution process begins. By checking the validity and completeness of configuration files upfront, the system avoids complex error handling during evolution execution. This preliminary action simplifies the overall system complexity by front-loading the validation work.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS11693861B2Systems and methods for providing data-driven evolution of arbitrary data structures

Publication Date: 2023.07.04 COGNIZANT TECHNOLOGY SOLUTIONS US CORP

AI summary

A system and process for generalizing an evolutionary process applied to a particular domain involving different problems includes a researcher module for generating a configuration specification applicable to a particular problem. An evolution module parses the configuration specification into a representative tree structure, assembles policies for each node in the tree structure, and generates candidate genomes using the policies for each node in the tree structure. The policies may be applied to new data or data from prior runs to generate candidate genomes. The evolution module translates internal representations of the generated candidate genomes into known representations of the candidate genome for evaluation in accordance with the particular domain parameters by a candidate evaluation module.