Subject:
|
Re: IR camera?
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Fri, 12 Feb 1999 19:01:53 GMT
|
Viewed:
|
1570 times
|
| |
|
|
lego-robotics@crynwr.com (Pete Hardie) writes:
> Laurentino Martins wrote:
> > I'm not an AI expert, so correct me if I'm wrong.
>
> You're wrong :->
>
> > To evolve, you need a purpose and some means to give it a score how it • performs in real life.
> > It would evolve into what? It needs a purpose. And why should it evolve in • the first place?
> > How would you say that "no, this design is no good". Grab the best of this • design and try to combine
> > it with the best of previous designs, adding some random data (I believe • this is called Genetic Algorithms).
> > And so on, and so on...
>
> All that is needed for evolution is variation and selection. Variation
> with in the 'genes' of
> the population would be the instructions to build each. Selection would
> be some method that
> would mark the 'unfit', so that they would not reproduce, or at least
> less well.
>
> Frex, let's say you have your factory, and each legobot carries its
> building instructions, and each
> is programmed to run a maze. You have the factory at the end, and it
> closes its doors after
> 10 legobots have entered. You will get fast maze running robots from
> this.
Don't be too sure. Unless you constrain it a lot you could get one or
two robots which travel fast enough to jam the door and the rest just meander
on through becuase it's jammed open. This is perhaps a bit contrived but I'd
be wary of making too specific statements of what a GA evolved system will do;
too many people have got caught on this before. An example is one of Karl Sims
creatures which was taking part in a GA which had maximim speed as the main
metric: it's strategy was just to grow tall and then fall over hence achieving
a maximum speed whilst falling over; I doubt if he would have been able to
predict this given the seemingly reasonable constraints he placed on it.
|
|
Message has 2 Replies:
 | | Genetic programming (was "Re: IR camera?")
|
| (...) Genetic programming in general often exhibits unintended results, namely: 1. Unless you're very careful, it ends up optimizing for the special cases rather than the general class of problem you want to solve. In the above example, you're (...) (25 years ago, 12-Feb-99, to lugnet.robotics)
| | | Re: IR camera?
|
| (...) True. However, in your case, you'd have evolved altruism. Still, the point was that 'purpose' is a perception overlaid on the system by humans. (25 years ago, 15-Feb-99, to lugnet.robotics)
|
Message is in Reply To:
| | Re: IR camera?
|
| (...) You're wrong :-> (...) All that is needed for evolution is variation and selection. Variation with in the 'genes' of the population would be the instructions to build each. Selection would be some method that would mark the 'unfit', so that (...) (25 years ago, 12-Feb-99, to lugnet.robotics)
|
17 Messages in This Thread:
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
This Message and its Replies on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
