"Evolutionary
Computation: The Road Behind, the Road Ahead"
Dr. Günter
Rudolph
|
Conference:
Abstract
The term Evolutionary Computation subsumes all those
randomized optimization methods whose design has been inspired by
principles of biological evolution. Since the appearance of
Evolutionary Algorithms (EAs) in the 1960s independently at different
places around the world, their protagonists had to grapple with taunt
and disrespect for a long time, especially in the academic world. The
situation was different in practice, when engineers were deploying EAs
with increasing frequency for finding good solutions for complex
optimization problems for which standard methods failed or were not
available.
This first part of the talk will trace the rocky road that Evolutionary
Algorithms have had to travel from the difficult beginning to the
comfortable present. Along the way I shall provide arguments supporting
my hypothesis that apart from the success stories in practical
applications the key to broad acceptance of Evolutionary Computation
was the development of a solid body of sound theory. A brief overview
of the main theoretical achievements will serve as the starting point
of the second part of the talk in which I shall outline my opinion
about future work, so that the field of
Evolutionary Computation keeps track on the road to a respected part of
science.