Re: [Toss-devel] Iterative deepening with pruning -- experiments

[Lukasz K. <luk...@gm...>]

> I agree, obviously. I'll add that I'd like to also do the idea of
> "action value priming" before weight learning. Everything going deeper
> should be after weight learning (because if weight learning works
> we'll be able to automatically learn a basic chess player -- features
> generated from rules should be enough.)

Yes, I think this is a good way. It is hard to speculate what
will work best and what not for me - let's test things and see.

Lukasz