Semantic Neutral Drift
Preprint
Abstract
We introduce the concept of Semantic Neutral Drift (SND) for evolutionary algorithms, where we exploit equivalence laws to design semantics preserving mutations guaranteed to preserve individuals' fitness scores. A number of digital circuit benchmark problems have been implemented with rule-based graph programs and empirically evaluated, demonstrating quantitative improvements in evolutionary performance. Analysis reveals that the benefits of the designed SND reside in more complex processes than simple growth of individuals, and that there are circumstances where it is beneficial to choose otherwise detrimental parameters for an evolutionary algorithm if that facilitates the inclusion of SND.Keywords
All Related Versions
This publication has 22 references indexed in Scilit:
- Probabilistic Graph Programs for Randomised and Evolutionary Algorithms2018
- Evolving Graphs by Graph Programming2018
- Neutral genetic drift: an investigation using Cartesian Genetic ProgrammingGenetic Programming and Evolvable Machines, 2015
- Cartesian Genetic Programming2011
- Finding needles in haystacks is harder with neutralityGenetic Programming and Evolvable Machines, 2006
- Finding Needles in Haystacks Is Not Hard with Neutrality2002
- The Advantages of Landscape Neutrality in Digital Circuit Evolution2000
- Cartesian Genetic Programming2000
- Through the labyrinth evolution finds a way: A silicon ridge1997
- Jungle evaluationLecture Notes in Computer Science, 1988