views:

192

answers:

1

Is it good idea to use genetic algorithm in production?

If you are using it: In what case? What pros for selecting subj? Can you easily add changes to algorithm?

+2  A: 

A typical scenario is to use GA as a search tool to find some 'innovative' design that improves some pre-existing man-made solution to given problems.

Once such a solution is found it can very well be used in 'production', or even mass production in the case of hardware manufacturing (just think that the first application ever of GA has been implemented as an optimization of design for aeronautic parts).

Also - talking about software that makes real-time use of GAs - in case of GA driven machine learning (there's a chapter on that on David E. Goldberg's GA bible), genetic algorithms will go 'live' and represent indeed the adaptive engine of the solver in question.

JohnIdol