Re: Bayesian Networks and Belief Functions

Joseph Halpern (halpern@cs.cornell.edu)
Mon, 7 Jun 1999 10:01:50 -0400 (EDT)

I think Uschi is completely right. To give a concrete example, if I
have a coin and I believe P(heads) = 0.5, my degree of belief that there
will be between 450,000 and 550,000 in the next million coin tosses is
practically 1. (I certainly would be prepared to bet large sums of
money on that event.) On the other hand, if I have no clue of the
probability of heads, then I also have no clue of the probabilty that
there will be between 450,000 and 550,000 heads in the next 1,000,000
coin tosses. -- Joe

-----------

As this is my first mail on this list: I am a PhD-student in Germany working
with/on Bayesian networks, and with a lot of interest in more or less
philosophical discussions on probability etc.

There is one point in the reasoning of David Poole that makes it not very
convincing to me: The reasoning depends on the question "What will be the
outcome of the next press?". This is not the only valid question.

> For those people who would like to distinguish ignorance for the outcome
> of a binary variables and probability 0.5, I would like to know how many
> different meanings are there to "I don't know" (for a binary random
> variables)?

As far as I can see it, there are as many different meanings as there are
different questions. As long as the question is, what the outcome of the
next press will be, the reasoning shows there is no difference between
ignorance and P(Q)=0.5. But if - for example - the question is about the
average in the next y presses, there is a difference. Thus, whether or not
it is important to make a distinction will be dependent on the question(s).

Uschi Sondhauss

--------------17EDB15DC10F5AF068187183