Re: Toaster to Generate Random Numbers

From: Tony T. Warnock (u091889@cic-mail.lanl.gov)
Date: 02/12/03 

From: "Tony T. Warnock" <u091889@cic-mail.lanl.gov>
Date: Wed, 12 Feb 2003 07:54:58 -0700

There is no bias. The excess number of 1's is N/Log(N) and thus as N
gets large the bias is 1/Log(N) which goes to zero.

The "strong" law of large numbers isn't very strong.

Dropping the leading 1's is in Postnikov's big article from the 1960s.
It removes the bias but not local clusters. (I think that there is no
way to remove those clusters if construction proceeds by concatination)

Relevant Pages

  • Re: Off-Topic -- Gender Discrimination
    ... What that graph is a prime example of how to lie with statistics.. ... assertions here which you provide no proof for. ... Bias is always guessing too high (or ... Ooops, it's zero. ...
    (rec.games.frp.dnd)
  • Re: equalizer error bias
    ... I have a channel with significant ISI with a 3 level ... The transmitted data and the receiver data after ... quantization have zero mean. ... I'm not sure why this would cause a bias. ...
    (comp.dsp)
  • The moving window or radar algorithm
    ... Hi I think I've found out how to make the bias element. ... processing of the two states one and zero. ... the circle of possible positions with a state of zero moving two steps ... Jaxon Modulation after being sub sampled to avoid runs of modulation ...
    (comp.compression)
  • Re: Actual photons?
    ... starting place aproaches zero as time goes on. ... You will find that the pointer gets farther and farther from the ... If you suspect a coin bias, flip twice as often and move the ... You can even do a computer simulation if you have a source of ...
    (sci.astro.amateur)
  • Re: Random
    ... no. Perez unbiasing depends on the amount of bias to be zero. ... > alter the sensitivity of the detector, ...
    (comp.security.misc)