Welcome to Bridge Zero Project.
I am bridge player and neural network developer (Python, Keras). Here I’m trying to build model that learns bridge bidding by itself i. e. without use of human knowledge. The purpose is to “generate” optimal bidding solution.
What does “optimal” mean? Optimal bid depends on system used by pair and surely optimal system depends on system used by opponents. So, by optimal system, I understand system that bids optimally against all possible systems (including itself).
It is possible that there is no minimax and there can exist systems that can beat optimal universal system.
Hi, this is a comment.
To get started with moderating, editing, and deleting comments, please visit the Comments screen in the dashboard.
Commenter avatars come from Gravatar.