Optimal nonlinear pricing by a regret‐averse monopoly