Business Analytics Exam 1 (OSU BUSMGT 2321)
define the problem
Decision Analysis step 1
list all alternatives (decision options)
Decision Analysis step 2
identify all possible states of nature (outcomes)
Decision Analysis step 3
identify payoffs (gain/loss from each combo of alternative and state of nature)
Decision Analysis step 4
choose modeling technique
Decision Analysis step 5
expected monetary value
EMV is an acronym for...
Expected Value of Perfect Information
EVPI
EVwPI - EMVmax
EVPI =
Expected Value with Perfect Information
EVwPI
optimistic
Maximax criteria is considered....
pessimistic
Maximin criteria is considered...
Monday morning Quarterback
Minimax Regret is considered as a ............................ strategy.
Equally likely criteria
a decision model where the probabilities of each state of nature are unknown, and the decision maker assumes the probability for each of the states are equal
infeasible solution
a solution to a linear program where NO points satisfy all of the constraints simultaneously
unbounded solution
a solution to a linear program where X1 and/or X2 are unconstrained; objective function value becomes more optimal as the variable approaches infinity
unique optimal solution
a solution to a linear program where only one of the corner points yields the optimal objective function value
multiple optimal solution
a solution to linear program where two corner yield the same, optimal objective function value (all other points on a line between these corner points are also optimal solutions)
EMV
a weighted average that is calculated by multiplying every payoff of a decision by the probabilities of each state of nature occurring
Equal Opportunity Loss
a weighted average that is calculated by multiplying every regret of a decision by the probabilities of each state of nature occurring
formulate
decision modeling step 1
solution
decision modeling step 2
interpret
decision modeling step 3
squares
decision tree models show decisions as .....................
circles
decision tree models show states of nature as ........................
operational (decisions)
decisions that take a limited amount of time and have a small consequence for an incorrect decision
strategic (decisions)
decisions that take a long time and has significant consequences for an incorrect decision (e.g. mergers/acquisitions)
tactical (decisions)
decisions that take a significant amount of time and can have a high consequence for the incorrect decision (e.g. forecasting production)
x1
horizontal axis variable
minimax regret criteria
identify maximum regret for each decision. then select the decision with the minimum regret.
Maximax Criteria
identify the maximum payoff for each decision. then select the decision with the maximum payoff
Maximin Criteria
identify the minimum payoff for each decision. then select the decision with the maximum payoff
constraints
limitations used in modeling to ensure the model conforms to reality
aggressive
maximax criteria indicates ....................... behavior
conservative
maximin criteria indicates ...................... behavior
competitive
minimax regret criteria indicates ..................... behavior
identify the best outcome for each state of nature
minimax regret table step 1
determine regret for each combination
minimax regret table step 2
identify max regret for each option
minimax regret table step 3
select decision with minimum regret (from maximums in step 3)
minimax regret table step 4
certainty
one of the four assumptions of linear decision modeling, all inputs to the model are correct/true
additivity
one of the four assumptions of linear decision modeling, all inputs to the model are cumulative (can be totaled)
proportionality
one of the four assumptions of linear decision modeling, all inputs to the model do NOT change regardless of production level
divisibility
one of the four assumptions of linear decision modeling, the optimal solution can have a decimal regardless if its impossible to produce a fraction of a unit
indifference point
quantity where decisions change
break-even point
quantity where π = 0
define, develop, & data
steps in formulate phase of decision modeling
analyze, sensitivity, & implement
steps in interpret phase of decision modeling
develop & test
steps in solution phase of decision modeling
shadow price
the change in the objective function value when the RHS of a constraint is increased by 1 unit
objective function
the equation you are trying to optimize (max./min.) in a model
certainty, proportionality, additivity, divisibility
the four assumptions of decision modeling with a linear program
decision tree
the model best suited to implement the concept of time
sensitivity
the optimal solution's ...................... is a measure of how much the objective function coefficients or RHS of the constraints can be changed before the optimal solution changes
corner points
the points where the constraints intersect to create the feasible region; the optimal solution to a linear program always occurs at one of these points
game theory
the science of logical decision making (particularly against a competitor)
x2
vertical axis variable