Recently, computational simulation has become a third approach-along with t
heory and laboratory simulation-to studying and solving scientific problems
. In this approach, a computer equipped with problem-solving software tools
may represent a virtual laboratory in which researchers can build a model
for a given problem and run it under varying conditions. These increasingly
complex computational methodologies require sophisticated models and techn
iques, and vice versa. The authors explain how developing and validating co
mplex models will increasingly depend on significant advances in experiment
al and testing techniques. High-performance parallel computers gave researc
hers the ability to implement inherently parallel techniques such as cellul
ar automata (CA), neural networks, and genetic algorithms-significant new m
athematical models for describing complex scientific phenomena.