P

Initializing...

The information budget of a k-valued optimal action is at most $\log k$ · Prove2Me