Engineering Decision Making and Risk Management

Chapter 5Decision Making Under Uncertainty

Learning Objectives:

After studying this chapter, the reader will be able to do the following:

1. Identify aleatory and epistemic uncertainties and describe the difference (Section 5.1).
2. Assess a subjective probability (Section 5.2).
3. Determine whether an alternative stochastically dominates another (Section 5.4).
4. Create a decision tree for a decision with uncertainty (Section 5.5).
5. Evaluate a decision tree and find the optimal policy (Sections 5.6 and 5.7).
6. Generate a valid utility function to represent a risk-averse decision-maker (Section 5.8).
7. Define different types of robustness measures (Section 5.9).
8. Identify ways to estimate uncertainty in the outputs of a model given uncertainty in the inputs (Section 5.10).
9. Use sensitivity analysis to propagate uncertainty (Section 5.10).
10. Use the method of moments to estimate the uncertainty of a model output (Section 5.11).
11. Use Monte Carlo simulation to estimate the uncertainty of a model output (Section 5.12).

Although life is full of uncertainties, it is unnecessary and impossible to consider all of them all of the time. The decision situations that were considered in the previous two chapters did not explicitly include any uncertainties. Although uncertainties were present, the evaluation of different attributes was done using expected values or most likely outcomes or somewhat vague or imprecise measurements.

When the uncertainties can have a significant impact on the outcome that ...