Øystein Schønning-Johansen wrote:
Sent: 12 December 2011 20:59
So, we don't care about the exactness of the absolute evaluation, we care about
the relative evaluation between the moves (or resulting positions after each
move). That is what makes it select good moves!
This strategy was
If gnubg could estimate the opponents rating based upon error rate then
it could be more aggressive
on a redouble since the other player is likely to make errors to modify
the cube decision.
This may have NOTHING to do with what your're talking about right now... :)
Tom
On 12/16/2011 06:43
thought about this
angle.
n Ian
From: Thomas A. Moulton [mailto:t...@moulton.us]
Sent: 16 December 2011 12:00
To: Ian Shaw
Cc: Øystein Schønning-Johansen; Mark Higgins; Frank Berger; bug-gnubg@gnu.org
Subject: Re: [Bug-gnubg] Neural network symmetry question
If gnubg could estimate
Subject: Re: [Bug-gnubg] Neural network symmetry question
If gnubg could estimate the opponents rating based upon error rate then it
could be more aggressive
on a redouble since the other player is likely to make errors to modify the
cube decision.
This may have NOTHING to do with what
I ran my test out to 400k runs and the symmetric network starting drifting
down, with the normal one edging it out in head to head competitions.
But the real evidence against it came from looking at probability estimates
from a couple benchmark positions: one where white is almost certainly
On Mon, Dec 12, 2011 at 9:11 PM, Mark Higgins migg...@gmail.com wrote:
I assume this is what the gnubg benchmark stuff is about btw? Comparing
probability estimates in a bunch of benchmark board positions against
rolled-out probabilities? How do you condense the many different cases into
a
I tried a little experiment on this: a 10-hidden-node network with a single
probability-of-win output, but two setups. The first doesn't have a whose turn
is it input and doesn't add any symmetry constraints. The second has the extra
inputs for the turn and makes the symmetry constraint I
My experience tells me that 100,000 trials may not be sufficient.
With today's computing power it should be easy to do at least a
couple of millions.
-Joseph
On 12 December 2011 11:22, Mark Higgins migg...@gmail.com wrote:
I tried a little experiment on this: a 10-hidden-node network with a
Thx - I'll run it longer and with more hidden nodes and see what happens.
On Dec 11, 2011, at 5:44 PM, Joseph Heled jhe...@gmail.com wrote:
My experience tells me that 100,000 trials may not be sufficient.
With today's computing power it should be easy to do at least a
couple of
You have an input that represents whose turn it is (one input for white, one
for black, value one if that player is on turn and zero otherwise). I think
that's in the original Tesauro setup isn't it?
On Dec 10, 2011, at 1:10 AM, Joseph Heled jhe...@gmail.com wrote:
Well, I am not sure how
Hi Mark,
If I take a given board and translate the position into the inputs and then
evaluate the network, it gives me a probability of win. If I then flip the
board's perspective (ie white vs black) and do the same, I get another
probability of win. Those two probabilities should sum to
Thx! Makes sense. Though I wonder if adding back in the whose move is it
input and reducing the hidden-output weights by half ends up as a net benefit
for training. Maybe I'll test it out.
On Dec 10, 2011, at 2:06 PM, Frank Berger fr...@bgblitz.com wrote:
Hi Mark,
If I take a given
I've been playing around a bit with neural networks for backgammon and found
something interesting, and want to see whether this is already part of gnubg.
Assume a Tesauro-style network with the usual inputs, and some number of hidden
nodes. And for simplicity, just one output representing the
Well, I am not sure how you flip the position, since it matters who is
on the move.
-Joseph
On 10 December 2011 16:17, Mark Higgins migg...@gmail.com wrote:
I've been playing around a bit with neural networks for backgammon and found
something interesting, and want to see whether this is
14 matches
Mail list logo
