Re: Representationalism rescues reinforcement learning


trial, error and (yawn) reinforcement creates internal model of the world that
can be created only by trial, error and (yawn) reinforcement

You forgot to account for the passage of time, a dimension of
unsurpassed utility. Yesterday's tribulation builds today's model that
wins tomorrow's trial.


I don't understand what you mean. Trial, error and reinforcement implies time dimension. You can't try without wasting some time. Every error is waste of time and every reinforcement algorithm has very important property: time complexity (how long does it take to learn something).

.

Relevant Pages

  • Re: Representationalism rescues reinforcement learning
    ... it is lucky enough to stumble on the appropriate response. ... error and (yawn) reinforcement" without building some ... you can't ceate internal model of the world if you don't learn by "trial, ...
    (comp.ai.philosophy)
  • Re: Representationalism rescues reinforcement learning
    ... trial, error and (yawn) reinforcement, is likely to get eaten before ... it is lucky enough to stumble on the appropriate response. ... you can't ceate internal model of the world if you don't learn by "trial, ...
    (comp.ai.philosophy)
  • Re: Representationalism rescues reinforcement learning
    ... it is lucky enough to stumble on the appropriate response. ... error and (yawn) reinforcement" without building some ... you can't ceate internal model of the world if you don't learn by "trial, ...
    (comp.ai.philosophy)
  • Re: Representationalism rescues reinforcement learning
    ... it is lucky enough to stumble on the appropriate response. ... you can't ceate internal model of the world if you don't learn by "trial, error and (yawn) reinforcement" ...
    (comp.ai.philosophy)