[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: computer-go: Zobrist hash random?

To: computer-go@xxxxxxxxxxxxxxxxx
Subject: Re: computer-go: Zobrist hash random?
From: Tapani Raiko <praiko@xxxxxxxxx>
Date: Fri, 13 Sep 2002 15:27:01 +0300 (EET DST)
In-reply-to: <7345.1031916591@xxxxxxxxxxxxxxxxx>
Reply-to: computer-go@xxxxxxxxxxxxxxxxx
Sender: owner-computer-go@xxxxxxxxxxxxxxxxx

> For example there should not be numbers like a,b,c where a xor b = c in my
> set of constants, for advoiding too early hash collisions.

Why go into the trouble? The probability of having one or more such
collisions in a random set is about 0.00000000001 (using 64 bits). There
is some serious research on e.g. random projections of high dimensional
data into lower dimensions for instance for text mining and I would not
call it nonsense.

--
 Tapani Raiko, <tapani.raiko@xxxxxxxxxxxxxxxxx>, +49 179 8927151,
 http://www.hut.fi/u/praiko

References:
- computer-go: Zobrist hash random?
  - From: nr997

Prev by Date: computer-go: Zobrist hash random?
Next by Date: Re: computer-go: Huima's hash function
Previous by thread: computer-go: Zobrist hash random?
Next by thread: Re: computer-go: Zobrist hash random?
Index(es):
- Date
- Thread