In order to tackle a problem using a GA, candidate solutions must be encoded in a suitable form. In the traditional GA solutions are represented by binary bit strings (`chromosomes'). While integer and decision variables are easily encoded in this form, the representation of continuous control variables is not so simple. In general, the only option available is to approximate them (rescaled as necessary) by equivalent integer variables. The accuracy with which an optimum solution can be resolved then depends on the encoded bit length of these integers, leading to an inevitable compromise between precision and execution time.
For combinatorial optimization problems, problem-specific solution encodings, such as ordered lists, are necessary. For example, a solution to the TSP can be represented by a string listing the cities in the order they are to be visited. Problem-specific operators are also required to manipulate such strings validly.