midterm exam - dsc-8240 murugan chellasamy. cases * cold fusion cold fusion * prescribed fire...
TRANSCRIPT
Midterm Exam - DSC-8240
Murugan Chellasamy
Cases
Cold Fusion Prescribed Fire GPC Strenlar
Cold Fusion
Objective Hierarchies Variables & Measures Problem Structuring
Consequence Table Influence Diagram Decision Tree
Mathematical Model - NA Solutions & Sensitivity
Cases
Cold Fusion - Objective Hierarchies
O b jec tive H ie ra rch y
M a x im ize E co n o m ica l E ffects M a x im ize P erso n a l S a tis fa ctio n
M a x im ize W ea lth /S elf S a tis fa ctio n
Cold Fusion - Variables & Measures
Decision Variables: Good for the Society Bad for the Society
Chance Variables Bill Passed Bill Rejected High Political Value Low Political Value Fusion Works Fusion Fails Economically Feasible Economically Infeasible
Consequence Actually Helps Society Does not make any difference
Cold Fusion - Consequence Table
Eco/Soc Effects Bill Political Value Fusion Feasibility
Good Passed High Works Feasible
Works Infeasible
Fails No Effect
Rejected Low No Effect No Effect
Bad Passed High Works Feasible
Works Infeasible
Fails No Effect
Rejected Low No Effect No Effect
Cold Fusion - Consequence Table
Cold Fusion - Influence Diagram
Good /BadFor
Society
High PoliticalValue
High/Low
FusionWorks/Fails
BillPassed/Rejected
High Satisfaction
Cold Fusion - Decision Tree
Cold Fusion - Solutions & Sensitivity
I believe the fate of this bill will be determined by the relative publicity it can generate to help the politicians and the scientists. It would be very hard to determine the economic and social benefits derived from Cold Fusion. It could be very beneficial to the society but if it meant that we would have to stop using the existing sources of energy, it may cause severe economic strains. On the other hand it may help the economy but may create problems to the society by making destructive technology available at a cheaper cost. Even if the bill gets passed, it may not be feasible to use this technology to reap the desired benefits.
Prescribed Fire
Objective Hierarchies Variables & Measures Problem Structuring
Consequence Table Influence Diagram Decision Tree
Mathematical Model Solutions & Sensitivity
Cases
Prescribed Fire - Objective Hierarchies
O b jec tive H ie ra rch y
A sso cia ted P ro b lem s
D ifferen t M eth o d s
C o st E ffic ien t B u rn in g
Prescribed Fire - Variables & Measures
Decision Variables: Burn Slash YUM
Chance Variables Controlled Fire Uncontrollable Fire - Escaped Fire Uncontrollable Fire - Problems
Consequence Cost Efficient Burn Escaped Fire Low Cost Burn Medium Cost Burn High Cost Burn
Prescribed Fire - Consequence Table
Method Burn Result Cost Amount
Slash Burn Controlled Minimal $85
Escaped Fire Substantial $1,000
Uncontrolled Low $125
Medium $200
High $500
YUM Controlled Minimal $185
Escaped Fire Substantial $1,000
Uncontrolled Low $225
Medium $300
High $600
Prescribed Fire - Per Acre
Prescribed Fire - Influence Diagram
Choose Method
ControlledFire
LowHigh
MediumCosts
UncontrolledFire
Cost Efficient Burn
Prescribed Fire - Decision Tree
Prescribed Fire Solutions & Sensitivity
Given the scenario’s, YUM seems to be the most cost efficient method .
GPC
Objective Hierarchies Variables & Measures Problem Structuring
Consequence Table Influence Diagram Decision Tree- NA
Mathematical Model Solutions & Sensitivity
Cases
GPC - Objective Hierarchies
O bje c tive H ie ra rc h y
S tu d y O p tio n s
C o n sid er D ifferen t S cen a rio s
E v a lu a te R isk /R etu rn
M a x im ize R etu rn s
GPC - Variables & Measures
Decision Variables: Product A Product B Product C
Chance Variables Delay in production Price Sales Market Share
Consequence Increased Market Share Lose Market Share Increased Return Reduced Margin
GPC - Consequence Table
Product Delay Price Sales Return (Millions)
A Yes High High Sales $5.00
Low Sales ($0.50)
Low High Sales $3.50
Low Sales $1.00
No High High Sales $8.00
Low Sales $0.00
Low High Sales $4.50
Low Sales $1.50
B Yes High $8.00
Medium $4.00
Low $0.00
C $1.00
GPC- Consequence Table
GPC - Influence Diagram
Choose Product
Delay Product
Sales
MaximizeReturn
GPC - Mathematical Model
Mathematical Models
EMVA delay = max (((0.3 * 5) + (0.7 * (-0.5))), ((0.5 * 3.5) + (0.5 * 1)))
EMVA no delay = max (((0.4 * 8) + (0.6 * 0)), ((0.5 * 4.5) + (0.5 * 1.5)))
EMVA = max (EMVA delay, EMVA no delay )
EMVB = (0.38 * 8) + (0.12 * 4) + (0.5 * 0)
EMVC = 1.0
EMV Optimal Decision = max( EMVA, EMVB, EMVC )
GPC - Solutions & Sensitivity
I was not able to get DPL to draw a decision tree. In any event, if we had more time and information we could have done more analysis. We could have used excel to set up a solver with constraints (upper and lower limits on probability) for products A & B to get a maximized return. We could also have set up a regression model with sales as the dependent variable with delay and price as independent variables to project profits with certain level of confidence. But, given the information we have, I would suggest we go with product B since we are a lot more certain about its out come. The decision will also be impacted by many other factors like the GPC’s financial strength and its position in the market etc.
Strenlar
Objective Hierarchies Variables & Measures Problem Structuring
Consequence Table Influence Diagram Decision Tree
Mathematical Model Solutions & Sensitivity
Cases
Strenlar - Objective Hierarchies
O b jectiv e H iera rch y
G a u ra n teed L ifes ty le S a fety
F in a n cia l/E m o tio n a lS ta b ility
O n e o f M a n y
S elf E steem P rid e
U n certa in ity
A g a in st T h e W o rld
M a x im ize S elf S a tis fa ctio n
Strenlar - Variables & Measures
Decision Variables: Accept Offer Reject Offer
Chance Variables Financial/ Physical/ Emotional State Peer & Social Pressure Future
Consequence Rich & Happy Broke & Miserable Self Esteem??
Strenlar - Consequence Table
Offer Lawsuit Process Fin State
Accept None Works Good
Fails Good
Reject Win Works Good
Fails Bad
Lose Works Bad
Fails Bad
Strenlar - Consequence Table
Strenlar - Influence Diagram
Accept/RejectOffer
Financial State
PhysicalState
EmotionalState
MaximizeSatisfaction
Strenlar - Decision Tree
Strenlar - Mathematical Model
The following diagram, done in Excel, accompanies the mathematical model used by Fred:
EMV Lawsuit = (0.6 * ((0.8 * 8,000) + (0.2 * (-200)))) + (0.4 * (-220))
EMVPI job = (0.8 * 2,443) + (0.2 * 343)
EMV lump sum = (0.8 * 1,252) + (0.2 * 500)
EMV Optimal decision = max ( EMV Lawsuit, EMVPI job, EMV lump sum )
Strenlar Solutions & Sensitivity
Fred’s position is unenviable. His decision will be influenced by how he feels about himself. I don’t think we can help him make a decision based on the financial outcomes. If Fred was a corporation, we could argue that he should reject the offer and take them to court since maximizing the value of the assets would be his only goal.