[Haskell-cafe] How would you hack it?

Andrew Coppin andrewcoppin at btinternet.com
Thu Jun 5 15:35:15 EDT 2008


Henning Thielemann wrote:
> "Markov chain" means, that you have a sequence of random experiments,
> where the outcome of each experiment depends exclusively on a fixed number
> (the level) of experiments immediately before the current one.
>   

Right. So a "Markov chain" is actually a technical way of describing 
something that's intuitively pretty obvious? (E.g., PPM compression 
works by assuming that the input data is some sort of Markov chain with 
as-yet unknown transition probabilities.)

> If the level is too high, you will just reproduce the training text.
>   

Yeah, I can see that happening! ;-)

The key, I think, is for the training set to be much larger than what 
you want to produce...



More information about the Haskell-Cafe mailing list