Introduction • Installation • Usage • Roadmap • Contributing • Authors • License
Evolutionary computing is one of the main techniques nowadays for solving complex optimization problems. This library provides with the basic structure for implementing the most common evolutionary algorithms, such as genetic algorithms.
Evolutionary algorithms basic structure
Evolutionary algorithms are composed basically by four elements:
- Individuals: Represent possible solutions of our problem in a determinate search space.
- Mutation: Mutation operator alterates one individual.
- Recombination: Recombination operator takes two parents and creates the offspring.
- Parent selection: Selection of the best parents that are going to be reproduced in the next generation.
- Survivor selection: Selection of the offspring and parents that are going to be the next generation.
This framework is going to provide the most common techniques for each component.
Currently project is under development (no stable version released
npm install genetics-js
No major versions have been released, so only Individuals creation is implemented:
import Genetics from 'genetics-js';
const { BinaryIndividual } = Genetics.individual;
let individual = new BinaryIndividual('001100');
individual.genotype // [false, false, true, true, false, false]The roadmap is strictly determined by the operations that are going to be implemented:
-
v0.1.0: Implementation of individuals. -
v0.2.0: Implementation of mutation operators. -
v0.3.0: Implementation of recombination operators. -
v0.4.0: Implementation of parent selection methods. -
v0.5.0: Implementation of survivor selection methods. -
v0.6.0: Implementation of population and offspring management. -
v0.7.0: Implementation of common evolutionary algorithms with fixed configurations.
You can report a bug, or request a feature with an issue:
Any help would be welcome 😄.
- Cristian Abrante - CristianAbrante - ([cristian@abrante.me])
This project is licensed under the MIT License