Epigenetic Gene Pool Control for Premature Convergence
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Evolutionary algorithms face challenges such as premature convergence and the inability to effectively utilize stepping stone individuals with critical parts due to fitness-based selection, leading to reduced diversity and potential loss of optimal solutions.
Innovation Solution
The implementation of an epigenetic mechanism that designates certain genetic material as inactive, allowing it to be preserved and reintroduced during procreation, thereby maintaining diversity and preventing premature convergence by giving 'free passes' to potentially valuable individuals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If fitness-based selection is used to select individuals for the next generation, then convergence speed is improved, but diversity is reduced leading to premature convergence
Solution Approach 1:
The system performs preliminary action by designating certain genetic material as epigenetic (inactive) before the selection process. This allows potentially valuable genetic material to be protected from fitness-based selection pressure, preserving diversity while still allowing convergence on fit individuals. The epigenetic designation acts as a pre-emptive measure to prevent loss of stepping stone individuals.
Solution Approach 2:
The system changes the state parameter of certain genetic material from active to inactive (epigenetic). This parameter change allows the genetic material to be preserved without participating in fitness-based selection, thereby maintaining diversity while allowing the active population to converge. The inactive state effectively removes the conflict between selection pressure and diversity preservation.
2Productivity
If stepping stone individuals with critical parts are discarded due to low fitness estimates, then selection efficiency is improved, but potential optimal solutions are lost
Solution Approach 1:
The system applies preliminary action by designating certain genetic material as epigenetic before fitness evaluation. This protects stepping stone individuals containing critical partial solutions from being discarded due to low overall fitness estimates. The epigenetic designation ensures that valuable genetic building blocks are preserved for future recombination, preventing information loss while maintaining efficient selection of fully fit individuals.
Solution Approach 2:
The system implements a modified discarding and recovering mechanism where genetic material can be temporarily set aside (discarded from active selection) as epigenetic, but then recovered and reintroduced into the active population through procreation. This allows stepping stone individuals to be effectively discarded from competition without permanent loss, as their genetic material can be recovered and combined with other individuals in future generations.
3Loss of time
If genetic material is actively selected based on fitness, then optimal solutions are found faster, but diversity maintenance becomes difficult
Solution Approach 1:
The system performs preliminary action by designating certain genetic material as epigenetic (inactive) before the fitness-based selection process begins. This creates a protected reservoir of genetic diversity that does not compete with active individuals. Meanwhile, active individuals can converge quickly on optimal solutions through fitness-based selection, while the epigenetic material serves as a diversity reservoir that can be reintroduced through procreation to prevent premature convergence.
Solution Approach 2:
The system introduces a new dimension to the population structure by adding the epigenetic (inactive) state alongside the active state. This dimensional expansion allows the system to simultaneously maintain diverse genetic material (in the epigenetic dimension) and efficiently select fit individuals (in the active dimension). The two dimensions operate semi-independently, resolving the time-diversity tradeoff by allowing both processes to occur simultaneously in different state spaces.
Data Source
AI summary
Roughly described, a computer-implemented evolutionary system evolves candidate solutions to provided problems. It includes a memory storing a candidate gene database containing active and epigenetic individuals; a gene pool processor which tests only active individuals on training data and updates their fitness estimates; a competition module which selects active individuals for discarding in dependence upon both their updated fitness estimate and their testing experience level; and a gene harvesting module providing for deployment selected ones of the individuals from the gene pool. The gene database has an experience layered elitist pool, and individuals compete only with other individuals in their same layer. Certain individuals are made epigenetic in the procreation module, after which they are not subjected to testing and competition. Epigenetic individuals are retained in the candidate gene pool regardless of their fitness. An epigenetic individual can become an active individual by random chance in procreation.