Autonomous Assessment of Demonstration Sufficiency via Bayesian Inverse Reinforcement Learning