Re: Bayesian boxes and expectation value

From: <GSLevy.domain.name.hidden>
Date: Fri, 21 May 1999 14:10:55 EDT

Gilles I fully agree with you. Your analysis is great.

In a message dated 99-05-20 04:06:41 EDT, you write:

<< If you don't know the maximum value M, there is actually no rational for
 switching. >>

I agree. Unless the maximum value of M is known there is no rational for
switching. If M is selected at random from the set of natural numbers, then
it is likely to be huge, in fact infinite.

A way to go around this problem is to limit M by selecting our test value m
from a ring say (A,B,C) with the property A>B, B>C, C>A ( as in the children
game of paper, stone [or hammer] and scissors.) Consider two boxes X and Y
each holding a random value in (A,B,C). After having selected box X and found
the value m inside, there is no way to know if box Y contains a better value.

Going back to the infinite set and refocussing on the MW topic the question
is how does this problem relates to the SSI thread?
Received on Fri May 21 1999 - 11:17:30 PDT