Evolutionary Computation

887 share

Evolutionary computation is a family of population-based trial and error functions that mimic biological evolution with a metaheuristic or stochastic approach.

While there are different techniques, evolutionary computation revolves around generating an initial set of candidate solutions and then updating them iteratively. This means the best solutions, as defined by a fitness test, get to “reproduce” a new generation. Usually involving by stochastically (randomly) removing poor fit solutions and introducing small random changes. This is intended to simulate natural selection and mutations to create a population of solutions with the best fit for the problem.

Common types of Evolutionary Computation Techniques:

• Swarm Intelligence

• Ant Colony Optimization

• Particle Swarm Optimization

• Evolutionary Algorithms

• Genetic Algorithms

• Gene Expression Programming