Module 2: Decisions Under Certainty

For some decisions, such as hardware selection, we are certain what the consequence of each possible choice will be, but it is not clear which choice is the best overall. In this session, we describe and compare several formal methods for handling such problems.

We start by taking a look at problem clarification, objective specification, and choice creation. We then study techniques that estimate the overall best choice. Our main focus will be on two methods:

• the even-swaps method,
• the Simple Multi-Attribute Rating Technique (SMART);

however, we will also mention two other approaches: SMARTER (SMART Exploiting Ranks) and AHP (the Analytic Hierarchy Process).

Case Study: The International Chernobyl Project found that SMART enabled people from different fields, such as scientists and government officials, to analyse the problem collectively and compare alternative strategies related to issues such as cancer deaths, monetary costs, and public acceptabilitly.
[Ref: S. French (1996), Multi-attribute decision support in the event of a nuclear accident, Journal of Multi-Criteria Decision Analysis, 5, 39-57]