Implementing variable environment genetic algorithms.

<p>(A) To promote a rule with to , we include extra neighbors keeping the same output with any combination of these two new neighbors. On the left we see a possible input and its respective output for a rule. On the right there are the respective possible outputs after generalization. Note that in the generalized rule the new inputs do not affect the rule's output. However, inter-generational mutation and crossover of the genetic material will yield changes in output that make the states of the new neighbors relevant. (B) Flowchart of the variable environment genetic algorithms. See the text for a more detailed description.