Classification of Cellular Evolutionary Algorithms

Cellular Evolutionary Algorithms (cEAs) are Evolutionary Algorithms (EAs) in which the population is structured with neighborhoods. An EA refers to the study of the foundations and applications of certain heuristic techniques based on the principles of natural evolution. It was 1960s that American and European researchers gave birth to stochastic search methods inspired by Darwinian evolution theory, all independently of each other. The most common Evolutionary Algorithms are [Huss98]:

• Genetic Algorithms (GAs) [Holl75]
• Genetic Programming (GP) [Koza92] [Koza94]
• Evolutionary Programming (EP) [FOW66]
• Evolution Strategies (ES) [Rech73]

In each case, the algorithm involves generating an initial population of individuals and then operating on that population iteratively so that over time the "quality" of the individuals persisting in the population tends to increase. The operations involved and the structure of the individuals in the population are what make the techniques different.

Though all three algorithms, EP, ESs, and GAs, originally have their distinguishing characteristics according to the level of organic evolution modelling, such classification is not useful anymore. For example, many GA practitioners have abandoned bit string for floating-point representations. ES practitioners often use crossover operators as a reproduction operator. Application of EP is no longer limited to the evolution of finite state machines. New evolutionary approaches such as genetic programming [Koza92] have been rising. Therefore, today all these systems are often collectively called Evolutionary Algorithm (EA) or Evolutionary Computation (EC).