Re: [Toss-devel] Monte Carlo gametree search (committed)

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

Note I haven't kept the "not today" promise. I've written the fitness
function and now I'm writing a random hillclimber for CNF formulas,
actually, a conjunction of existentially and universally quantified
cnfs: "(ex x y. cnf1) and (all x y. cnf2)"

On Mon, Sep 28, 2009 at 10:06 PM, Lukasz Kaiser <luk...@gm...> wrote:
>> I've committed the MC gametree search code.
>
> This is good news - it seems that most of the technical part
> is now there :). I can also report that it works for me (well,
> at least compiles and "make tests" works). Great work!
>
> Lukasz
>