Dynamic Stochastic Learning in Regulatory Optimization: Deadlines and Error with Continuous and Discrete Evidence