The expected value of perfect information in the optimal evaluation of stochastic systems | lit.salon