Home MadLab   Further development
Further development
 

As of now, the MadLab algorithm is quite poweful as long as the problem does not have too large lambda-order, and as long as the problem does not have to be decomposed into sub-problems.

After recently having spent much time on the user interface of MadLab, the plan is to concentrate fully on the engine/algorithm, in order to attack some harder problems. More specifically, the goal for 2004/2005 is to have MadLab solve the following problem in reasonable time (it actually can solve it, but it consumes way too much time and memory).

White to capture the five black stones at "a"

The position is from a game between Takagawa Kaku (B) and Go Seigen (W). The black block can be captured, but there is, among other things, a counter-attack on the five white stones to the left of the black block to consider. Hopefully much more on this when MadLab understands how to decompose such problems!