decision making
TRANSCRIPT
Decision MakingDecision Making
Sales – Price Relationship AnalysisSales – Price Relationship Analysis
A workshop making lampshades finds that the number it can sell varies
depending on selling price.
It can sell 10 per week if the price is set at Rs. 80, but 50 per week if the
price is reduced to Rs. 40. The cost of production is Rs. 20 for each
lampshade and there are overheads of Rs. 60 per week.
Assume linear relationship between price and sales.
What should be the price to maximize his profit.
Sales – Price Relationship AnalysisSales – Price Relationship Analysis
0
500
1000
1500
2000
2500
0 20 40 60 80
Sales
Ru
pe
es
Revenue Cost Profit
Inverse CostsInverse Costs
Maintenance department of a foundry wants to plan its annual expenditure on
equipment maintenance.
Currently it has a crew of 10 people. It costs the company Rs. 20000 per
month per crew member.
If department increases its crew size, it can make maintenance operations
more efficient. As a result breakdown costs will come down.
Data analysis showed that size of maintenance crew and breakdown loss
have a inverse relation as follows.
Inverse CostsInverse Costs
Crew size 10 11 12 13 14 15
Ependiture 2400000 2640000 2880000 3120000 3360000 3600000
Breakdown loss 12000000 6000000 4000000 3000000 2400000 2000000
Total cost 14400000 8640000 6880000 6120000 5760000 5600000
Crew size 16 17 18 19 20
Ependiture 3840000 4080000 4320000 4560000 4800000
Breakdown loss 1900000 1800000 1700000 1600000 1500000
Total cost 5740000 5880000 6020000 6160000 6300000
Inverse CostsInverse Costs
0
2000000
4000000
6000000
8000000
10000000
12000000
14000000
16000000
8 9 10 11 12 13 14 15 16 17 18 19 20 21
Crew size
Co
st
Expenditure Breakdown loss Total cost
Inventory CostsInventory Costs
An
nu
al c
os
tA
nn
ual
co
st
Lot Size (Lot Size (QQ))
Holding cost (Holding cost (HCHC))
Ordering cost (Ordering cost (OCOC))
Total cost = Total cost = HCHC + + OCOC
Replacement DecisionsReplacement Decisions
DepreciationDepreciation
Any equipment we use at work reduces in value year by year,
which is called as depreciation.
Calculation of depreciation is needed for many decision making
situations and one of them is replacement analysis.
There are two basic methods of depreciation calculation.
1. Straight line analysis
2. Declining Balance method
Depreciation and Replacement Depreciation and Replacement AnalysisAnalysis
A machine tool costs Rs. 300,000 when new.
Lets calculate the written down value after 1,2 and 3 years using
1. Straight line method with annual depreciation of Rs. 50,000
2. By declining Balance method with annual depreciation of 20 %
Approach 1: Straight line method
Capital cost = 3,00,000
Annual depreciation = 50,000
Value after 1st year = 2,50,000
Value after 2nd year = 2,00,000
Value after 3rd year = 1,50,000
Depreciation and Replacement Depreciation and Replacement AnalysisAnalysis
A machine tool costs Rs. 300,000 when new.
Lets calculate the written down value after 1,2 and 3 years using
1. Straight line method with annual depreciation of Rs. 50,000
2. By declining Balance method with annual depreciation of 20 %
Approach 2: Declining balance method
Capital cost = 3,00,000
Annual depreciation = 20%
Value after 1st year = 30,00,000 - 0.2 x 3,00,000 = 2,40,000
Value after 2nd year = 2,40,000 – 0.2 x 2,40,000 = 1,92,000
Value after 3rd year = 1,92,000 – 0.2 x 1,92,000 = 1,53,600
Equipment Replacement DecisionsEquipment Replacement Decisions
Suppose a factory has a permanent need for an equipment, that wears
out over a period of several years. In the initial period of use, the
depreciation is likely to be high but maintenance costs will be low.
Towards the end of its useful life, the rate of depreciation may be slow
but maintenance costs will be high.
When will it be better to sell off the existing equipment and purchase a
new one ?
Equipment Replacement DecisionsEquipment Replacement Decisions
When will it be better to sell off the existing equipment and purchase a new one ?
Year Depreciationmaintenance
Cost1 50000 60002 45000 75003 40000 120004 35000 200005 30000 340006 25000 500007 20000 700008 15000 90000
Equipment Replacement DecisionsEquipment Replacement Decisions
0
20000
40000
60000
80000
100000
120000
0 2 4 6 8 10Year
Ru
pe
es
Depreciation Maintenence Total Cost
Decision making Under UncertaintyDecision making Under Uncertainty
A set of quantitative decision-making techniques for decision situations
where uncertainty exists
States of nature
events that may occur in the future
decision maker is uncertain which state of nature will occur
decision maker has no control over the states of nature
Decision making Under UncertaintyDecision making Under Uncertainty
Payoff TablePayoff Table
A method of organizing & illustrating the payoffs from different decisions
given various states of nature
A payoff is the outcome of the decision
Payoff TablePayoff Table
States Of Nature
Decision a b
1 Payoff 1a Payoff 1b
2 Payoff 2a Payoff 2b
Decision making CriteriaDecision making Criteria
Maximax criterion (optimistic)
choose decision with the maximum of the maximum payoffs
Maximin criterion (Pessimist)
choose decision with the maximum of the minimum payoffs
Minimax regret criterion
choose decision with the minimum of the maximum regrets for
each alternative
Hurwicz criterion
choose decision in which decision payoffs are weighted by a
coefficient of optimism,
coefficient of optimism () is a measure of a decision maker’s
optimism, from 0 (completely pessimistic) to 1 (completely optimistic)
Equal likelihood (Laplace) criterion
choose decision in which each state of nature is weighted equally
Decision making CriteriaDecision making Criteria
A B C DX 8 0 -10 6Y -4 12 18 -2Z 14 6 0 8
Pay-Offs in Thousands of rupeesAlternative
X -10 8Y -4 18Z 0 14
Alternative Minimum Pay-off Maximum Pay-off
Decision Making ExampleDecision Making Example
A B C DX 8 0 -10 6Y -4 12 18 -2Z 14 6 0 8
Pay-Offs in Thousands of rupeesAlternative
X -10 8Y -4 18Z 0 14
Alternative Minimum Pay-off Maximum Pay-off
Maximin Maximax
Decision Making ExampleDecision Making Example
Minimax Regret ExampleMinimax Regret Example
A B C
S1 700 300 150S2 500 450 200S3 300 300 100
Events and Pay-offsStrategic Altenatives
A B C
S1 0 150 50S2 200 0 0S3 400 150 100
Strategic Altenatives
Events and Regrets
Minimax Regret ExampleMinimax Regret Example
A B C
S1 700 300 150S2 500 450 200S3 300 300 100
Events and Pay-offsStrategic Altenatives
A B C
S1 0 150 50 150S2 200 0 0 200S3 400 150 100 400
Maximum Regret
Strategic Altenatives
Events and Regrets
Minimax Regret ExampleMinimax Regret Example
A B C
S1 700 300 150S2 500 450 200S3 300 300 100
Events and Pay-offsStrategic Altenatives
A B C
S1 0 150 50 150S2 200 0 0 200S3 400 150 100 400
Maximum Regret
Strategic Altenatives
Events and Regrets
Hurwicz CriterionHurwicz Criterion
Step 1: Choose alfa and (1-alfa)
Step 2: Determine for each alternative,
h = (alfa) (max pay off) + (1-alfa) (minimum pay off)
Step 3: Select the alternative with maximum value of ‘h’
‘alfa’ is the coefficient of optimism. It is a measure of a decision maker’s optimism, from 0 to 1 (completely optimistic)
(1-alfa) is the degree of pessimism
Hurwicz Criterion ExampleHurwicz Criterion Example
Take degree of optimism as 0.6
A B CS1 8000 4500 2000S2 3500 4500 5000S3 5000 5000 4000
Strategic Altenativ
Events and Pay-offs
For alternative S1, h = 0.6(8000)+0.4(2000) = 5600
Hurwicz Criterion ExampleHurwicz Criterion Example
Take degree of optimism as 0.6
A B CS1 8000 4500 2000 5600S2 3500 4500 5000 4400S3 5000 5000 4000 4600
Strategic Altenative
Events and Pay-offsh
Hurwicz Criterion ExampleHurwicz Criterion Example
Take degree of optimism as 0.6
A B CS1 8000 4500 2000 5600S2 3500 4500 5000 4400S3 5000 5000 4000 4600
Strategic Altenativ
Events and Pay-offsh
Laplace Criterion ExampleLaplace Criterion Example
In this method we each state of nature is weighted equally.
In other words, probability of occurrence of events is considered to be equal.
Step 1: Assign equal weights to each pay off of an alternative or strategy.
Step 2: Estimate the expected pay off for each alternative
Step 3: Select the alternative which has the maximum expected pay off
Laplace Criterion ExampleLaplace Criterion Example
A B C D
1 4 0 -5 32 -2 6 9 13 7 3 2 4
Events and Pay offsAlternative
Expected Pay off for Alternative 1:
0.25 (4) + 0.25 (0) +0.25 (-5) + 0.25 (3) = 0.5
Laplace Criterion ExampleLaplace Criterion Example
A B C D
1 4 0 -5 3 0.52 -2 6 9 1 3.53 7 3 2 4 4.0
Events and Pay offsAlternative
Expected Pay off
Expected Pay off for Alternative 1:
0.25 (4) + 0.25 (0) +0.25 (-5) + 0.25 (3) = 0.5
Laplace Criterion ExampleLaplace Criterion Example
A B C D
1 4 0 -5 3 0.52 -2 6 9 1 3.53 7 3 2 4 4.0
Events and Pay offsAlternative
Expected Pay off
Expected Pay off for Alternative 1:
0.25 (4) + 0.25 (0) +0.25 (-5) + 0.25 (3) = 0.5
Decision making With ProbabilitiesDecision making With Probabilities
Decision making With ProbabilitiesDecision making With Probabilities
Probabilities need to be assigned to events
Expected value is a weighted average of decision outcomes.
EV x p ix ixi
n
where
ix outcome i
p ix probability of outco
( )
1
me i
Expected Monetary Value CriterionExpected Monetary Value Criterion
A store keeper stocks a perishable item. Shelf life of this item is one month. Store keeper wants to determine the number of items he should stock at the beginning of the month.
He buys the item for Rs. 30 and sells at Rs. 50.
He analyzes the trend for last two years i.e. 24 months. The following table gives the sales during last 24 months.
Sales 10 11 12 13
Frequency 3 5 10 6
Expected Monetary Value CriterionExpected Monetary Value Criterion
Sales 10 11 12 13
Frequency 3 5 10 6
Probability 0.125 0.208 0.417 0.250
Expected Monetary Value CriterionExpected Monetary Value Criterion
10 11 12 13
10 200 170 140 11011 200 220 190 16012 200 220 240 21013 200 220 240 260
StockDemand
Expected Monetary Value CriterionExpected Monetary Value Criterion
10 11 12 13
10 25.00 21.25 17.50 13.7511 41.67 45.83 39.58 33.3312 83.33 91.67 100.00 87.5013 50.00 55.00 60.00 65.00
EMV 200.00 213.75 217.08 199.58
Stock and conditional pay offDemand
Expected Monetary Value CriterionExpected Monetary Value Criterion
10 11 12 13
10 25.00 21.25 17.50 13.7511 41.67 45.83 39.58 33.3312 83.33 91.67 100.00 87.5013 50.00 55.00 60.00 65.00
EMV 200.00 213.75 217.08 199.58
Stock and conditional pay offDemand
Expected Regret CriterionExpected Regret Criterion
10 11 12 13
10 0 30 60 9011 20 0 30 6012 40 20 0 3013 60 40 20 0
Stock and RegretDemand
10 11 12 13
10 0.00 3.75 7.50 11.2511 4.17 0.00 6.25 12.5012 16.67 8.33 0.00 12.5013 15.00 10.00 5.00 0.00ER 35.83 22.08 18.75 36.25
Stock and Conditional RegretDemand
Decision TreesDecision Trees
Decision TreesDecision Trees
Bharat Oil Company (BOC) owns a land that may contain oil.
Geologist report shows a 25% chance of oil
Another company is offering to buy the land for Rs. 90 Cr
If BOC decides to drill, it will earn a profit of Rs. 700 Cr if oil is found.
However, it will incur a loss of Rs. 100 Cr if oil is not found.
Should BOC drill or sell ?
Decision TreesDecision Trees
Oil
Oil
Dry
Dry
(0.25)
(0.25)
(0.75)
(0.75)
700 Cr
-100 Cr
90 Cr
90 Cr
Drill
Sell
decision
Expected pay off is 100 Cr
Expected pay off is
90 Cr
Value of Perfect InformationValue of Perfect Information
In many decision making exercises it is possible to get more or extra information about the events or state of nature.
But it will cost extra money.
Question : Is additional information worth the cost ?
Value of Perfect InformationValue of Perfect Information
Continuing with the previous example,
A sesmic survey can tell whether the land is fairly likely or fairly unlikely to have oil.
Cost of the survey is Rs. 30 Cr
Should BOC do the survey ?
Value of Perfect InformationValue of Perfect Information
Expected pay off with perfect information is
= 0.25 (700) + 0.75 (90) = 242.5 Cr
Expected value of perfect information is
= Expected pay off with perfect information - Expected pay off without perfect information
= 242.5 – 100 = 142.5 Cr.
If EVPI is less than the cost of survey, then don’t do the survey. It’s not worth it.
In this case, 142.5 Cr. >> 30 Cr.
It is worthwhile doing the survey.
Decision TreesDecision Trees
A firm is adding a new product line and must build a new plant. Demand will
either be favourable or unfavourable, with probabilities of 0.6 and 0.4,
respectively. If a large plant is built and demand is favourable the pay off is
estimated to be Rs. 1520 Cr. If the demand is unfavourable, the loss with larger
plant will be Rs. 20 Cr
If a medium sized plant is built and demand is unfavourable, the pay off is Rs.
760 Cr. If the demand proves to be favourable, the firm can maintain the medium
sized facility or expand it. Maintaining medium sized facility will result in to a pay
off of Rs. 950 Cr and expanding it will give a pay off of Rs 570 Cr.
Draw a decision tree for this problem
What should the management do to achieve the highest expected pay off ?
Decision TreesDecision Trees
Fav
Un Fav
(0.6)
(0.6)
(0.4)
(0.4)
1520 Cr
-20 Cr
760 Cr
Large
Small
decision Fav
Un Fav
Expand
Continue
570 Cr
950 Cr
0.6 (1520) – 0.4 (20) = 904 Cr
0.6 (950) + 0.4 (760) = 874 Cr
Build a large Plant