HW-6: Decision-Making under Risk
Due 10/18
CSCI 510, Fall 2017
30 points
The MedFRS project has selected Company L to develop, produce, and field test the First Responder Workstations (FRWs), and train the first-responder personnel in their use.
Company L has identified two candidate approaches for developing and producing the FRWs. The Bold approach B will give them a profit of $800K if it succeeds, but a loss of $200K if it fails. The Conservative Approach C will give them a profit of $400K for sure.
- (12 points). Initial analysis
- Construct the payoff matrix for the decision situation.
- Determine the best decisions and payoffs under:
- the Maximin rule;
- the Maximax rule; and
- the Laplace rule.
- Determine the breakpoint probability of Bold approach success above which the Bold approach will be preferable.
- (12 points). Value of perfect information
- Determine the Expected Value of Perfect Information (EV)perfect info: the payoff if Company L had perfect information about the success or failure of the Bold approach. Assume that the original success or failure probabilities were equally likely at 0.5 each.
- Determine the Expected Value of No Information (EV)no info: the payoff if Company L had no information about the success or failure of the Bold approach. Assume that the original success or failure probabilities were equally likely at 0.5 each.
- Determine the Expected Value of Acquiring Perfect Information (EVPI): the maximum investment in acquiring information on the success of the Bold approach that would be worth doing.
- (6 points). RRL for prototypes
Company L also identifies two levels of prototyping that would buy information to reduce the risk of proceeding with no information. Prototype X would cost $30K and reduce the cost of B failing from $200K to $100K. Prototype Y would cost $50K and reduce the cost of B failing from $200K to $20K.
- Compute the Risk Reduction Leverage (RRL) for Prototypes X and Y.
- Identify which is the better investment.