Minimax Theorem
A cornerstone result in game theory and statistical decision theory, asserting that in a class of games satisfying convexity and compactness, minimizing the maximum loss and maximizing the minimum gain are equivalent.
Timeline (last 90 days)
For each restriction condition available to the bettor, the authors identify, via the minimax theorem over finite-support distributions of answers, the corresponding sharp label-free lower bound on calibration residual.
The minimax theorem was first proved by von Neumann in 1928, asserting that in a class of games satisfying convexity and compactness, minimax and maximin are equivalent.
All Facts (2)
For each restriction condition available to the bettor, the authors identify, via the minimax theorem over finite-support distributions of answers, the corresponding sharp label-free lower bound on calibration residual.
50%UnverifiedThe minimax theorem was first proved by von Neumann in 1928, asserting that in a class of games satisfying convexity and compactness, minimax and maximin are equivalent.
50%