Re: f(x) and NN

["P.J.Leonard" <P.J.Leonard@xxxxxxxxxx>]

Weimin Xiao wrote:
> 
> Hi, there,
> 
> I would think P.J. Leonard's position = f(situation) function goes a little
> bit too far. A simpler expression can be
>     value = f(board)
> which is what we know as an 'influence function'. Current influence
> functions as oriented from Zoborist's diamond are simple minded. The effort
> of using simple minded influence functions makes GO NP hard (such as min-max
> search). The effort of reducing NP calculation at somewhat fraction yields
> knowledge representation, pattern match, and all the AI and code twisting.
> 
> The point: we only need a good evaluation function. "Mathematical Go
> Endgames", Benson's rule, etc. are not easy to use but have limited ability.
> How to write the evaluation function? I would suggest the neural network is
> a simple way to approach the function if we do not expect the mathematical
> beauty.



I think you need both 

position = function(situation)  AND  value = function(situation).

I call these functions 

List<Point> moveSuggest(boardState)
Value       evaluate(boardState)


 These functions are used in the tree search.

moveSuggest()    adds children 
evaluate()       prunes tree and directs search towards higher values.



if moveSuggest() is perfect you do not need evaluate()
if  evaluate() is perfect you can suggest all possible moves and do a 1
ply search.


I think it is highly unlikely that we will ever see perfect functions.


-- 
 Cheers Paul 
           
 ``Reality is for people who lack imagination''