Components in a Genetic Algorithm

A population of solutions is created and each member of it is tested to determine how good it is. If there is no solution in the population whose quality is acceptable, the solutions are manipulated by picking the better ones and modifying them to form a fresh population whose members are tested in turn. This cyclic process continues over many generations, during which the population improves until, one hopes, a solution of suitable quality emerges. The entire process, from the preparation of the first population to the final identification of a solution of near-optimum quality, makes up a GA run. 

A method to represent potential solutions to a problem as individuals that can be manipulated by evolutionary operators. 

A recipe that can be used to assess the quality of individuals. 

A method to select the better individuals to act as parents in the creation of a new generation. 

A mechanism for genetic change, so that a newly created member of the population need not be an exact copy of a previous member. 

---