
122 ◾ Biological Computation
• Solution representation: Each solution will be represented as a com-
bination of the elements specied above as expression trees, such as
the trees shown in Figure3.7.
• Initial population: As with other evolutionary algorithms, the com-
putation begins with a random population of candidate solutions. To
generate this population we need to generate a set of random expres-
sion trees. e maximal depth of the trees is usually specied in
advance and guides the process. In the full method, randomly cho-
sen operator nodes are added to the tree successively until the maxi-
mum depth is reached, and beyond that only ...