Evolutionary Algorithm Pool Segmentation for Diversity

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Evolutionary algorithms often face premature convergence to local optima and discard stepping stone individuals based solely on fitness estimates, leading to a lack of diversity and potential loss of optimal solutions in data mining environments.

Innovation Solution

A data mining system that promotes individual diversity by using a candidate individual pool with a competition module that assigns individuals to experience layers based on their testing experience level and partition tags, allowing competition only within similar groups, thereby ensuring a diverse set of patterns and preventing early elimination of stepping stone individuals.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If individuals are discarded based solely on fitness estimates, then the selection process becomes simple and efficient, but diversity is lost and stepping stone individuals are prematurely eliminated

Engineering Contradiction:

Improveselection efficiencyVSAvoidindividual diversity

Core Design Contradiction:

ProductivityVSQuantity of substance

Solution Approach 1:

The candidate individual pool is segmented into multiple experience layers based on testing experience levels. Each layer contains individuals with similar experience levels, allowing differentiated competition strategies. This segmentation prevents premature elimination of stepping stone individuals by giving them competition only within their peer group rather than against all individuals regardless of experience level.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Different competition rules are applied to different experience layers. Individuals in each layer compete according to their specific experience level characteristics. This local quality approach ensures that stepping stone individuals are evaluated fairly against peers with similar experience, preserving diversity while maintaining selection efficiency within each local group.

Inventive Principle:

Principle #3Local quality

2Speed

If all individuals compete in a single pool, then competition is maximized and convergence is accelerated, but stepping stone individuals are lost and optimal solutions may be missed

Engineering Contradiction:

Improveconvergence speedVSAvoidsolution optimality

Core Design Contradiction:

SpeedVSReliability

Solution Approach 1:

The single competition pool is segmented into multiple experience layers, each with its own competition dynamics. This allows parallel evolution in different layers, maintaining solution diversity while still achieving convergence within each layer. The segmentation prevents premature loss of stepping stone individuals while maintaining overall convergence progress.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The competition structure is extended from a single-dimensional pool to a multi-dimensional layered structure. Individuals are organized not just by fitness but by experience level, creating an additional dimension for competition. This dimensional expansion allows stepping stone individuals to survive in appropriate layers while still enabling overall system convergence through inter-layer progression.

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

3Quantity of substance

If experience layers are introduced to preserve diversity, then stepping stone individuals survive longer, but the system complexity increases

Engineering Contradiction:

Improvepattern diversityVSAvoidsystem architecture

Core Design Contradiction:

Quantity of substanceVSDevice complexity

Solution Approach 1:

The system is segmented into experience layers with clear boundaries and progression rules. Each layer has defined entry and exit criteria based on testing experience levels, making the complexity manageable through modular organization. This segmentation preserves diversity while containing system complexity through structured layer management.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system uses experience level as a dynamic parameter to determine competition group assignment. As individuals gain testing experience, their experience level parameter changes, automatically moving them between layers. This parameter-based approach manages complexity through automated, rule-based transitions rather than manual intervention.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS11288579B2Training and control system for evolving solutions to data-intensive problems using nested experience-layered individual pool

Publication Date: 2022.03.29 COGNIZANT TECHNOLOGY SOLUTIONS US CORP

AI summary

Roughly described, in an evolutionary technique for finding optimal solutions to a provided problem, a computer system uses a grouping algorithm that is better able to find diverse and optimum solutions in data mining environment with multiple solution landscapes and a plurality of candidate individuals. Each candidate individual identifies with a potential solution, and is associated with a testing experience level and one or more partition tags. Each candidate individual is assigned into one of a plurality of competition groups in dependence upon the individual's testing experience level and partition tag. During competition among candidate individuals, a candidate individual can only replace another candidate individual if both the candidate individuals have a common partition tag and are in the same competition group. A candidate individual cannot replace another candidate individual if they have different partition tags or are in different competition groups.