Re: What did that thread indicate?


Traveler <traveler@xxxxxxxxxx> wrote:

>On 15 Sep 2005 15:53:17 -0700, humiguel@xxxxxxx wrote:
>
>>A reinforcement-learning machine can, at least in principle,
>>generate secondary behaviour by combining primary behaviours,
>>testing that in the environment and learning from the feedback.
>
>I agree but I would not use the word 'generate'. Reinforcement
>learning is a selection mechanism. It does not generate anything. It
>assumes the a priori existence of behaviors.

Obviously.

>It is used for combining
>low-level behaviors into high-level apetitive and aversive behaviors.

What about neutral behaviours?

>A different mechanism and principle (motor coordination) is
>responsible for actually generating/creating the behaviors.

I find this business of creating secondary behaviour the most
difficult aspect of AI, right next to dealing with the huge amount
of stimuli in a natural environment. For this latter aspect,
concepts constitute a good abstraction mechanism and are very
efficient a reducing the complexity of the world.

However, for creating behaviour there's not a clear-cut mechanism
that matches the power of concepts for the input part. So the
important question that needs to be answered is: What are the
abstraction mechanisms appropriate for encoding behaviour?


Antonio Esteves

--
Corby - A new approach to Artificial Intelligence
http://futalgo.planetaclix.pt/corby/index.htm

.

Relevant Pages

  • Re: Reinforce learn this
    ... controller, which works as well for the robot, that our brain works for us. ... give us a few years of exposure to an environment with ... Only some basic learning hardware was there at ... This is where reinforcement learning comes in. ...
    (comp.ai.philosophy)
  • Re: The Brain
    ... aka the correct behavior to produce, then the learning ... It has no creativity. ... be entirely due to the environment. ... it must learn to apply that reward to other future situations ...
    (comp.ai.philosophy)
  • Re: Robotics, AI, and Ethics
    ... giving it negative rewards for illegal moves. ... is not how you make the learning algorithm stronger. ... the learner not in the environment. ...
    (comp.ai.philosophy)
  • Re: Robotics, AI, and Ethics
    ... because the one you wrote to play TTT isn't ... I think the "strength" of a learning system is limited ... And if the only reward ... environment we are dealing with, ...
    (comp.ai.philosophy)
  • Re: Is the human brain a optimal generic learning system?
    ... to fit whatever environment they are placed in. ... It can be an optimizer (no generics involved) that ... playing chess, ... And when a programmer is in the learning ...
    (comp.ai.philosophy)
Loading