No Text.
No Text.
I think that it is correct as Genetic. Incidentally, genetic programming is a very cool topic.
Genetic algorithms are methods for choosing the best way to deal with data based on fitness functions. In essence, it's trying to model evolution in computer algorithms. The idea is simple -- you start with a "population" of data, from which you want to select the best to fit a certain task. You encode each of the individuals somehow (e.g. as a bitstring). Now, you need a way to choose some random individuals to see whether you're getting closer to the solution. The fitness function determines how "good" the given selection is -- you're trying to get the best result possible.
You also have to encode it so that the individuals mutate randomly (helps introduce fresh perspectives that can help to solve the problem), e.g. by very occasionally flipping a bit of a bitstring. Generally, you want the GA to get closer to the solution by keeping good matches around. Repeat and rinse.
GAs have the benefit of often avoiding local minima since they can change a lot within a short space of time (so there's no need for simulated annealing here IIRC).
Genetic algorithms usually give good, robust solutions, but can be beaten by domain-specific specialised methods.
Genetic programming is where you take a soup of program instructions/operands (sine, cosine, +, /, =, 0, 1, 2 and so on), and attempt to evolve a program! You create parse trees randomly, mutate them by crossing over the branches, and see what happens. There have been a few nifty takes on this. Of course, you can define domain-specific operations to chuck into the programming.
I can't remember what it's called, but there is definitely something cool out there where a guy created a mini world somewhere for viruses and other 'organisms', in the simplest way possible. It produced neat results such as parasites (damn, I'll have to try to find that link again).
A good resource is The Hitch-Hiker's Guide to Evolutionary Computing
"All paid jobs absorb and degrade the mind."
<br />-- Aristotle
You can certainly delete this post, I was just curious if it was a typo.
EAP
Ah, I remember its name now. It was Tierra.
And there's no point in deleting a thread like this, really! I'd prefer to make it possible for newbies to stumble across info in the most unexpected places, wouldn't you?
Besides, I'm sure this won't be the last time someone thinks Genetic = Generic...
"All paid jobs absorb and degrade the mind."
<br />-- Aristotle
Bookmarks