THEORY OF DECISIONS PHASE 2 - SOLVE PROBLEMS BY APPLYING THE ALGORITHMS OF THE UNIT 1 DELIVERY OF THE ACTIVITY
Aldemar González González Martínez. Code 1099365901 Group 212066_31
September 2018 National open and distance university (UNAD). CEAD Valledupar.
Introduction
The successful decision-making is a high impact issue for the successful progress of any of the social, economic, political and educational approaches, where you can always expect to generate the best result, for this the importance of tools such as the decision tree, According to the evaluation and study carried out, the results of this study lead us to determine which of the options is the most useful for what I am looking for. Knowing how to use the tool means that we can find the expected value of the perfect information and the expected value of the information show, variables that determine with figures that are close to reality and thus be able to guide a decision maker towards the best alternative , helping to avoid risks.
DEVELOPMENT OF THE ACTIVITY
Problem 1. DECISION TREES, EVPI and EVMI
Teratex, a textile company that has a productive experience in the foreign market of 25 years, must decide if it manufactures a new product in its main plant, or if on the contrary the purchase from an external supplier. The profits depend on the demand of the product. The table shows projected profits, in millions of dollars. Table 1. Decision process for the commercialization of the product States of nature Decision Demand low- Demand low Demand High alternative utility average - utility utility Manufacture 221 251 310 Subcontract 210 225 278 Buy 195 236 289 Probabilities Ʃ =
1
0,35
0,42
0,23
a. Use EVPI to determine if the company should try to get a better estimate of the demand. DEMANDA Manufacture
Node 1
Subcontract
Node 2
Node 3
254,07 Buy
EVPI EVwith PI EVwithoutPI EVPI
Node 4
Low
0,35
221
Half
0,42
251
High
0,23
310
Low
0,35
210
Half
0,42
225
High
0,23
278
Low
0,35
195
Half
0,42
236
High
0,23
289
254,07
231,94
233,84
N O I S I C E D 1 E D O N T L U S E R
It would be recommended to
Node 2 manufacture with an expected value of $ 254,07 254.07 million for being the one that
EVwith PI - EVwithout PI 254,07 254,07 0,00
The expected value of the perfect information is 0
Manufacture
High
0,45
310
Node 4
Half
0,39
251
272,62
Low
0,16
221
0,425 FAVORABLE
Node 2
Subcontract
Buy
High
0,45
278
Node 5
Half
0,39
225
246,34
Low
0,16
210
High
0,45
289
Node 6
Half
0,39
236
253,17
Low
0,16
195
Node 1 MARKET RESEARCH
264,69
Manufacture
UNFAVORABLE
Node 3
Subcontract
0,575
Buy
Decision nodes chosen
High
0,28
310
Node 7
Half
0,44
251
258,82
Low
0,28
221
High
0,28
278
Node 8
Half
0,44
225
235,40
Low
0,28
210
High
0,28
289
Node 9
Half
0,44
236
239,06
Low
0,28
195
Interpretation: the market study would have a payment of 264.64 million, If favorable, the best option would be Fabricate with a payment of 272.62 million; equally if it were unfavorable, the best option would also be Fabricate with a payment of 258.82 million. c. What is the expected value of market research information? EVMI EVMI EVMI
EVwhithMI - EVwithout PI 264,69
254, 07
10,62
d. What is the efficiency of the information? E
EVMI (10,62) EVPI (0,0)
0,00
0%
Interpretación: the market study would be 100% efficient because the EVPI is 0
Problem 2. DECISION TREES, EVPI and EVMI
ElectroCom, a company that manufactures electronic components for the introduction in its product catalog, must decide whether to manufacture a new product in its main plant, subcontract it with company supervision or if it buys it from an external supplier. The profits depend on the demand of the product. The table shows projected profits, in millions of dollars. Table 2. Decision process for the commercialization of the product States of nature Decision Demand low- Demand low Demand High alternative utility average - utility Medium - utility Manufacture Subcontract Buy Lease Outsource
173 181 183 125 188
183 192 197 128 192
195 207 207 131 198
Demand High utility 218 213 215 137 209
0,19
0,21
0,28
0,32
Probabilities Ʃ
=1
a. Use EVPI to determine if t he company should try to ge t a bett er est imate of the demand. Low Manufacture
Subcontract
Node 1
Buy
Node 2
Node 3
Node 4
202,9
Lease
Outsource
EVPI EVwith PI EVwithoutPI EVPI
Node 5
Node 6
0,19
DEMANDA 173 183 195,66 195 218
Low Average
0,21
High Medium High
0,28 0,32
Low
0,19
Low Average
0,21
192
High Medium
0,28
207
High
0,32
213
Low
0,19
183
Low Average
0,21
197
High Medium High
0,28 0,32
207 215
Low
0,19
125
Low Average High Medium
0,21 0,28
128 131
High
0,32
137
Low
0,19
188
Low Average High Medium
0,21 0,28
192 198
High
0,32
209
181
200,83
202,9
N O I S I C E D 1 E D O N T L U S E R
It would be recommended to Buy with Node 4 an expected value of $ 202,9 million for 202,9 being the one that generates the best utility
131,15
198,36
EVwith PI - EV without PI 204,81 202,9 1,91
Interpretación: The expected value of the perfect information is 1,91
b. A test market study of potential product demand is expected to re port a favorable (F) or unfavorable (U) condition. The relevant FAVORABLE Conditional
Pre vious probabilities
Low
0,19
0,2
0,04
0,11
Low Average High Medium High
0,21 0,28
0,2 0,35 0,5
0,04 0,10
0,12 0,29 0,47
Probabilities
0,32
P (F)
Joint Probabilities
Later
Sta te
0,16 0,338
Probabilities
UNFAVORABLE Sta te
Pre vious probabilities
Low
0,19 0,21 0,28
Low Average High Medium High
Node 4
201,92
Subcontract
Node 5
205,05 0,338 FAVORABLE
Buy
Node 6
206,85
Subcontract
Node 7
132,79
Buy
Node 8
201,34 MARKET RESEARCH
0,8 0,8 0,65 0,5 P (U)
0,32
Manufacture
Node 2
Conditional Probabilities
Joint Probabilities
0,15 0,17 0,18
Later Probabilities
0,23
0,16 0,662
Low Low Average High Medium High
0,11 0,12 0,29 0,47
173 183 195 218
Low Low Average High Medium High
0,11 0,12 0,29 0,47
181 192 207 213
Low Low Average High Medium High
0,11 0,12 0,29 0,47
183 197 207 215
Low Low Average High Medium High
0,11 0,12 0,29 0,47
125 128 131 137
Low Low Average High Medium High
0,11 0,12 0,29 0,47
188 192 198 209
Low Low Average High Medium High
0,23 0,25 0,27 0,24
173 183 195 218
Low Low Average High Medium
0,23 0,25 0,27
181 192 207
High
0,24
213
Low Low Average High Medium
0,23 0,25 0,27
183 197 207
High
0,24
215
Low Low Average High Medium
0,23 0,25 0,27
125 128 131
High
0,24
137
Low Low Average High Medium
0,23 0,25 0,27
188 192 198
High
0,24
209
0,25 0,27 0,24
Node 1
202,90
Manufacture
Node 9
192,46
Subcontract
Node 10
198,67
UNFAVORABLE
Buy Node 3
0,662
Node 11
200,89
Decision nodes chosen Subcontract
Node 12
130,31
Buy
Node 13
196,84
Interpretation: the market study would have a payment of 202.90 million, if favorable, the best option would be to buy with a payment of 206.85 million; equally, if it were unfavorable, the best option would also be Buy with a payment of 200.89 million.
c. What is the expected value of market research information? EVMI EVMI EVMI
EVwhithMI - EVwithoutPI 202,90
204,81
-1,91
d. What is the efficiency of the information? E
EVMI (-1,91) EVPI (1,91)
*
100
-100%
Interpretation: It is observed that the EVMI is not efficient against the EVPI, so it would not be considered necessary to carry out a market study, since instead of helping, it generates a 100% inefficiency.
Problem 3. DECISION TREES, EVPI and EVMI
Teratextyl, a textile company that has a productive experience in the foreign market of 30 years, must decide if it manufactures a new product in its main plant, or if on the contrary the purchase from an external supplier. The profits depend on the demand of the product. The table shows projected profits, in millions of dollars. Table 3. Decision process for the commercialization of the product States of nature Decision alternative
Demand low-
Demand low average -
Demand High
Demand High -
utility
utility
Medium - util ity
utility
Manufacture
85
87
91
95
Subcontract
78
81
85
89
Buy
82
85
87
90
Lease
83
85
87
91
Outsource
85
87
89
93
0,30
0,22
0,25
0,23
Probabilities Ʃ
=1
a. Use EVPI to determine if the company should try to get a better estimate of the demand. Low Manufacture
Subcontract
Node 1
Buy
Node 2
Node 3
Node 4
89,24
Lease
Outsource
EVPI EVwith PI EVwithoutPI EVPI
Node 5
Node 6
0,30
Low Average
0,22
High Medium
0,25
High
0,23
DEMANDA 85 87 89,24 91 95
Low
0,30
78
Low Average
0,22
81
High Medium
0,25
85
High
0,23
89
Low
0,30
82
Low Average
0,22
85
High Medium
0,25
87
High
0,23
90
Low
0,30
83
Low Average
0,22
85
High Medium
0,25
87
High
0,23
91
Low
0,30
85
Low Average
0,22
87
High Medium
0,25
89
High
0,23
93
82,94
85,75
N O I S I C E D 1 E D O N T L U S E R
It would be recommended to Node 2 Manufacture with an expected value of 89,24 $ 89,24 million for being the one that generates the best utility
86,28
88,28
EVwith PI - EVwithout PI 89,24 89,24 0
Interpretation: The expected value of perfect information is 0
b. A t est market s tudy of potential product demand is expected to report a favorable (F) or unfavorable (U) condition. The relevant FAVORABLE Sta te
Previous proba bilitie s
Low Low Average High Medium High
0,22 0,35 0,33
Conditional Probabilities
0,07 0,08 0,08
0,30 0,22 0,25 0,23
0,42
Joint Probabilities
P (F)
0,10 0,322
Later Probabilities
0,20 0,24 0,26 0,30
UNFAVORABLE Sta te
Previous proba bilitie s
Low Low Average
0,78 0,65 0,67
High Medium High
0,58
Manufacture
Node 4
90,01
Subcontract
Node 5
83,81 0,322 FAVORABLE
Node 2
Buy
Node 6
86,40
Subcontract
Node 7
86,90
Buy
Node 8
88,90 MARKET RESEARCH
Conditional Probabilities
0,30 0,22 0,25 0,23 P (U)
Joint Probabilities
0,23 0,14 0,17
Later Probabilities
0,35 0,21
0,13 0,678
Low Low Average High Medium High
0,20 0,24 0,26 0,30
85 87 91 95
Low Low Average High Medium High
0,20 0,24 0,26 0,30
78 81 85 89
Low Low Average High Medium High
0,20 0,24 0,26 0,30
82 85 87 90
Low Low Average High Medium High
0,20 0,24 0,26 0,30
83 85 87 91
Low Low Average High Medium High
0,20 0,24 0,26 0,30
85 87 89 93
Low Low Average High Medium High
0,35 0,21 0,25 0,20
85 87 91 95
Low Low Average High Medium High
0,35 0,21 0,25 0,20
78 81 85 89
Low Low Average High Medium High
0,35 0,21 0,25 0,20
82 85 87 90
Low Low Average
0,35 0,21
83 85
High Medium High
0,25 0,20
87 91
Low Low Average High Medium High
0,35 0,21 0,25 0,20
85 87 89 93
0,25 0,20
Node 1
89,24 Manufacture
Node 9
88,87
Subcontract
Node 10
82,53
UNFAVORABLE
Node 3
Buy
0,678
Node 11
85,44
Decision nodes chosen Subcontract
Node 12
85,98
Buy
Node 13
87,98
Interpretation: the market study would have a payment of 89.24 million, If favorable, the best option would be Fabricate with a payment of 90.01 million; equally if it were unfavorable, the best option would also be Fabricate with a payment of 88.87 million.
c. What is the expected value of market research information? EVMI EVMI EVMI
EVwhithMI - EVwithoutPI 89,24
89,24
0
d. What is the efficiency of the information? E
EVMI (0) EVPI (0)
*
100
0%
Interpretation: There is NOT something concrete, since the EVPI and EVMI give me 0, for which neither of the two options is pertinent to choose.
SCREENSHOTS OF THE PRACTICAL STAGE
Exercise 1
Exercise 2
Exercise 3
Conclusions.
This phase 2, has contributed in a positive way in the way of how to face the decisions under premises of uncertainty and risk, which allows to speed up the work of professionals through the use of useful tools such as the decision tree associated with the expected value of the perfect information and expected value of the sample information, determining with accuracy the degree of efficiency between the one and the other, in such a way that the decision making is given in a more complete and accurate way.
In the same way, this phase has been a demanding and competitive challenge so that as a student I can learn to develop and undertake new knowledge that strengthens my quality as a professional, where discipline and entrepreneurship helped to understand the issues in a consistent and necessary way for what is required.
Bibliography according to APA standards
Sanderson, C. (2006). Analytical Models for Decision Making. New York, USA: McGraw-Hill Education Editorial. Retrieved from http://bibliotecavirtual.unad.edu.co:2051/login.aspx?direct=true&db=nlebk&AN=23 4098&lang=es&site=eds-live Gilboa, I. (2001). A Theory of Case-Based Decisions. Camdridge, UK: Cambridge University Press Editorial. Retrieved from http://bibliotecavirtual.unad.edu.co:2051/login.aspx?direct=true&db=nlebk&AN=72 982&lang=es&site=eds-live Rokach, L. (2008). Data Mining With Decision Trees: Theory And Applications , Bern, Switzerland: H. Bunke, University Bern, Switzerland. Retrieved from http://bibliotecavirtual.unad.edu.co:2051/login.aspx?direct=true&db=nlebk&AN=23 6037&lang=es&site=eds-live
