Optimal Reporting of Predictions

摘要: Abstract Consider a problem in which you and group of other experts must report your individual predictive distributions for an observable random variable X to some decision maker. Suppose that the each expert is assigned prior weight by maker these weights are then updated based on observed value X. In this situation will try maximize updated, or posterior, appropriately choosing distribution report, rather than necessarily simply reporting honest distribution. We study optimal strategies under various conditions regarding knowledge beliefs about reports experts, utility functions posterior weight. present only it always Attention restricted problems can take finite number values.