Re: [Bug-gnubg] New positions for training database

[Ian Shaw]

Frank Berger wrote:

> > Interesting. That's a surprise to me. I would have thought there would
> > be benefits. I haven't played much nackgammon myself, so I don't know
> > how games evolve. Maybe with both sides having four men back, reasonably
> > timed one-sided backgames rarely evolve.
> For GNU it may work better, because of the different training methods 
> (BGBlitz uses straight TD(lambda)-Training). I would guess it's worth 
> a try.
In contrast, Gnubg used TD training to get to intermediate level, then 
supervised training to get to it's current expert level after TD 
training settled at a certain level. I had assumed that TD can only 
reach a certain limit. It's interesting that bgblitz has achieved the 
same standard as gnubg through TD alone.

Has anyone tried going back to TD training after a period of supervised 
training? Perhaps now that gnubg has got over its sticking point it can 
learn more from TD. Perhaps there are subtleties it can now discover now 
that it is already an expert.
> > Another approach might start from prototype holding games and backgames
> > such as found in Kit Woolsey's Encyclopaedia. The idea being that these
> > are common positions, so it is useful to have plenty of data for
> > accurate training.
> If you enter the positions, could you please send me a copy?
Certainly, but it won't be anytime soon, if I do it at all.

-- Ian