Population-based incremental learning
In machine learning and soft computing, Population-based incremental learning (PBIL) is a type of genetic algorithm where the genotype of an entire population is evolved rather than individual members[1].
Genotype representation
In PBIL, genes are represented as real values in the range [0,1], indicating the probability that any particular allele appears in that gene.
Algorithm
The PBIL algorithm is as follows:
- A population is generated.
- The fitness of each member is evaluated and ranked.
- Update population genotype based on fittest individual.
- Repeat steps 2-3
References
- ^ Karray, Fakhreddine O.; de Silva, Clarence (2004), Soft computing and intelligent systems design, Addison Wesley, ISBN 0-321-11617-8