Learning by Doing—An Adaptive Approach to Multiperiod Decisions

Publisher Website
Google Scholar
Abstract
A decision-maker is required to make a decision in each period over a sequence of periods. His decision in any period will affect: (1) the immediate consequence resulting from his decision in that period, and (2) the new information based on which decisions in future periods will be made. An adaptive approach to the above multiperiod decision process is suggested here, in which a decision-maker acts while learning from his experience so as to maximize the expected accumulative utility payoff from the entire decision process. It is demonstrated that the value of information, in this situation, depends not only on (1) the uncertainty of environment, (2) the nature of information, (3) the utility payoff function, but also on (4) the expected rate of learning.

This publication has 0 references indexed in Scilit: