Computational stopping

Decision making processes usually involve time-consuming activities of collecting, processing and communicating information and computing and implementing the final decisions. Typically, if the environment does not change, the effectiveness of the final decisions improves with the amount of time spent on these activities. On the other hand, if the environment is changing with time, the final decisions will become more obsolete with longer delays, because the conditions for which they were designed will have changed. The problem of choosing optimal delays by balancing improvement against obsolescence is treated in this paper. Upon providing motivational examples the problem is formulated and illustrated, first in the single stage and then in the dynamic programming framework in two different ways. An economically interesting and informationally efficient optimal stopping policy is indicated.