V(Si, A2)

V(S2, A2)

JpV(S1, A2) + (1 -p)V(S2, A2)

In this table, we have explicitly laid out the states of nature, the probability of their occurrence and the actions. Each entry in the last column is the value of a particular action given the state of nature. We are then able to compute the average values of different actions; for example, the average value of action A1 will be ^>V(S1, A1) + (1 -^)V(S2, A1). The hard work, of course, is filling in the entries in the table. For example, to estimate the probability of the alternative states of nature we can use methods of likelihood theory or Bayesian analysis. To evaluate the effects of alternative actions with uncertainty, we can use Monte Carlo forward simulations.

0 0

Post a comment