Re: Non-Random Coin Flip?
From: "Douglas A. Gwyn" <DAGwyn@null.net>
Date: Sat, 12 Apr 2003 00:15:43 GMT
> So in that case a file full of all "1" could qualify as 'random'.
There are two kinds of randomness that have formal definitions and
are generally accepted among workers in the field. One of them,
associated with Chaitin among others, allows a string of data to
be rated in terms of how big a computer program would have to be
to generate that string. However, the more usual notion of
randomness pertains to processes, not to their outputs, and using
that approach, it is a mistake to try to attribute randomness to
a specific sample of data.
- Re: Kolmogorov Complexity - again
... becomes less and less relevant as the string continues to ... between the notions of randomness and inductive inference. ... Kolmogorov, Solomonoff, Von Mises, Chaitin, etc. all had the same ... it is quite clear that the choice of UTM only matters in the ...
- Re: does sqrt(2) exist in CM?
... > If the two senses of randomness were the same, then the string ... Wow, Jesse, now you've unraveled a huge mystery. ... > | same probability. ...
... >>> When you are talking KC of a string, ... >>> compressed without regard to its function, but this compression will ... >>> maintain the original function of a string may require that the string ... >>> deterministic system ends up having the appearance of true randomness. ...
... >>> erm, if your string is minimum, then it is compressed for the reference ... >> compressed without regard to its function, but this compression will ... the original function is destroyed. ... >> deterministic system ends up having the appearance of true randomness. ...
- Re: random infinite strings
... for sure the string is not random. ... is a test that tells that an infinite string is _not_ random, ... martin-lof randomness coincides with chatin randomness ... If you know the seed and the RNG, you can predict the outcoming process. ...