Re: linear programming for soln of games in extensive form
[ QUOTE ]
Does anyone have experience with this? I am familiar with the normal form LP, but the extensive form formulation ala Koller et. al. isn't clicking yet. I can't quite figure out how to formulate the constraints, or what they are in fact supposed to represent.
I can give more details, but will wait for "yeah, I've done that" before going to the trouble.
eastbay
[/ QUOTE ]
I've done quite a bit of this.
Get your game in the form
max_x min_y x^T H y
s.t.
E x = e
F y = f
x,y >= 0
and then converting this to a standard LP is easy.
What bit is causing you grief?
Marv
|