Dinusha Dissanayake

Jan 15, 2021

3 min read

Nature Inspired Optimization Algorithms — Genetic Algorithm

This is the part 4of the article series Nature Inspired Algorithms. Introductory article could be found at https://dindisn.medium.com/nature-inspired-optimization-algorithms-part-1-introduction-11dafcba07db

Genetic Algorithm is an evaluation based algorithm which follows the abstract process of Charles Darwins Evaluation theory and natural selection In biological systems.It has come into light that most of the companies at present use these as a back support system for the combinatorial optimization problems in their routine optimization problems like scheduling, planning and in many more scenarios.

Natural behaviour

Darwins theory starts from an existing population. Individuals in the population become parents and produce off springs and through heredity, characteristics of parents gets passed to the children. Variations will be introduced to the population through mutations happening in some scenarios.

Generally, as population tends to over produce only the fittest will survive and through natural selection some will survive whilst the others won’t. In natural scenarios this could be seen where the taller giraffes survive more than the shorter ones where there are only tall trees to consume as food, darker moths survive in industrially polluted blackened environment where as lighter ones are easily seen an captured to predators. So the traits of theses survivors will be more prominent in the survived and produced populations in future populations and this is how the evaluation happens.

Technical aspects

In the genetic algorithm, concepts like natural selection, crossover, mutation have been mirrored.