Re: which distribution ?
- From: John Mauer <jmauer@xxxxxxxxxx>
- Date: Fri, 24 Feb 2006 07:30:47 EST
Try the Logarithmic distribution. It can give the very peaked response at 1 while also providing large values in the tail.
p(x) = theta^x/(x*ln(1-theta))
.
- Follow-Ups:
- Re: which distribution ?
- From: Eric B
- Re: which distribution ?
- References:
- which distribution ?
- From: Eric B
- which distribution ?
- Prev by Date: Re: Pattern matching with statistics
- Next by Date: Re: Describing linear densities and patterns
- Previous by thread: Re: which distribution ?
- Next by thread: Re: which distribution ?
- Index(es):
