Evolutionary Computation: An Overview

Evolutionary computation is an area of computer science that uses ideas from biological evolution to solve computational problems. Many such problems require searching through a huge space of possibilities for solutions, such as among a vast number of possible hardware circuit layouts for a configuration that produces desired behavior, for a set of equations that will predict the ups and downs of a financial market, or for a collection of rules that will control a robot as it navigates its environment. Such computational problems often require a system to be adaptive— that is, to continue to perform well in a changing environment. Problems like these require complex solutions that are usually difficult for human programmers to devise. Artificial intelligence practitioners once believed that it would be straightforward to encode the rules that would confer intelligence on a program; expert systems were one result of this early optimism. Nowadays, however, many researchers believe that the “rules” underlying intelligence are too complex for scientists to encode by hand in a top-down fashion. Instead they believe that the best route to artificial intelligence and other difficult computational problems is through a bottom-up paradigm in which humans write only very simple rules and provide a means for the system to adapt. Complex behaviors such as intelligence will emerge from the parallel application and interaction of these rules. Neural networks are one example of this philosophy; evolutionary computation is another. Biological evolution is an appealing source of inspiration for addressing difficult computational problems. Evolution is, in effect, a method of searching among an enormous number of possibilities—e.g., the set of possible gene sequences—for “solutions” that allow organisms to survive and reproduce in their environments. Evolution can also be seen as a method for adapting to changing environments. And, viewed from a high level, the “rules” of evolution are remarkably simple: Species evolve by means of random variation (via mutation, recombination, and