Decisionmaking rule-based

Bandit design. An adaptive allocation design based on ideas of sequential decisionmaking with the following features. (1) Results to date are used to determine which treatment a patient will receive. (2) The patient does not necessarily receive whatever treatment is currently believed to be the best because it is recognized that there may be value in studying less-studied treatments. (3) The allocation rule attempts to maximize the (discounted) totality of future patients on the better treatment (including the one about to be treated). 

Sun and Yih (1996) adopted their idea to develop a neural network-based controller for manufacturing cells. In their approach, a neural network was trained to serve as decisionmaker that will select a proper dispatching rule for its associated machine to process the next job. Based on their results, the controller performs well under multiple criterion environments. In addition, when the production objectives change, the controller can respond to such change in a short time. 

The problem of dike-heightening is similar to the problem of maintenance optimization the imcertain occurrence of failures affects the planning and timing of preventive maintenance actions. When using a risk-based decision rule, the optimal time between repairs is often determined by the interval with the lowest present value of the costs involved. Due to the probabilistic nature of such an approach, this value is uncertain and the expectation is used for decisionmaking. However, practical experience has shown that the variance of this optimum is usually quite large see, for example. Van Noortwijk (2003). In these cases, it is desirable to choose a maintenance interval which is less optimal, but of which the outcome is less uncertain. en making such a decision one is often said to be risk averse. 

---