UNIVERSIDAD NACIONAL ABIERTA Y A DISTANCIA - UNAD

Escuela de Ciencias Básicas Tecnología e Ingeniería - ECBTI

TEORIA DE LAS DECISIONES - (212066A_951)

 

 
Unit 1 - Task 2 - Solving of decision models under risks
 

 
Produced by
Jairo A. Guacaneme
John E. Escobar
Jorge L. Montoya
Juan C. Garcia
Sebastian López
 

Group: 23
 

 
Tutor
Andrea Del Pilar Lopez
 
 

 
 
Universidad Nacional Abierta y a Distancia
School of Basic Sciences Technology and Engineering
Industrial engineering
March/14/2021
 Unit 1 - Task 2 - Solving of decision mode

Exercise 1. DECISION TREES, EVPI an

A textile company that has a productive experience in the foreign market of 25 years, must decide if it manufactures
the purchase from an external supplier. The profits depend on the demand of the product. The table show

Table 1. Decision process for the commercialization of the product

States
Demandof nature
low Demand
Decision alternative Demand average - High -
low-utility utility utility
Manufacture 721 822 678
Subcontract 735 718 756
Buy 679 919 817
Probabilities Ʃ = 1 0.43 0.37 0.2

According to the corresponding information in Table 1 and the Predicted Value of Perfect Information (EVPI) theory, the Expe
Trees, respond:

a. Use EVPI to determine if the company should try to get a better estimate of the demand.

b. A test market study of potential product demand is expected to report a favorable (F) or unfavorable (U) condition. The rele

P(F/low)= 0.4 P(D/low) = 0.6

P(F/low P(D/ low
average)= 0.3 average) = 0.7
P(F/high) = 0.45 P(D/high) = 0.55

c. What is the expected value of market research information?

d. What is the efficiency of the information?

Probability

low-utility 0.43

Manufacture low average 0.37

Nodo 2
749.77
High - utility 0.2

low-utility 0.43

Nodo 1 Subcontract low average 0.37

Nodo 3
(decision) 732.91
 High - utility 0.2

low-utility 0.43

Buy low average 0.37

Nodo 4
795.4
High - utility 0.2

The expected value without perfect information (VESiP) indicates that the decision to make is to buy with a value expected of

VEIP Predicted Value of perfect information

VEIP= Valor absoluto (VEcIP - VEsIP)
VEIP= 24.08 Answer: The expected value perfect information (VEIP) Or demand of the t

FAVORABLE
PREVIOUS
PROBABILITIES P(Ai)
es la probabilidad CONDITIONAL JOINT PROBABILITIES P(FՈAi) se
previa del estado de la PROBABILITIES P(F/Ai) Es calcula teniendo en cuenta la regla de
STATE OF NATURE (Ai) la probabilidad de que
naturaleza la multiplicacion para eventos
sea favorable dado el dependientes
estado de la naturaleza
(Ai)

Sumatoria igual a 1 Sumatoria igual a la P(F)

P(F/low)= 0.43 0.4 0.172

P(F/low average)= 0.37 0.3 0.111
P(F/high) = 0.2 0.45 0.09
1 P(F) 0.373

UNFAVORABLE
PREVIOUS
PROBABILITIES P(Ai) CONDITIONAL
PROBABILITIES P(U/Ai) JOINT PROBABILITIES P(UՈAi) se
es la probabilidad calcula teniendo en cuenta la regla de
STATE OF NATURE (Ai) previa del estado de la Es la probabilidad de que la multiplicacion para eventos
naturaleza sea favorable dado el
estado de la naturaleza dependientes
(Ai)

Sumatoria igual a 1 Sumatoria igual a la P(F)

P(F/low)= 0.43 0.6 0.258

P(F/low average)= 0.37 0.7 0.259
P(F/high) = 0.2 0.55 0.11
1 P(F) 0.627
 PROBABILITY

low-utility

Manufacture low average

Nodo 4
740.680965
High - utility

low-utility
FAVORABLE
0.373 Nodo 2 Subcontract low average
Nodo 5
(decision) 735.008043
High - utility

low-utility

Buy low average

Nodo 6
783.718499
High - utility

Nodo 1 PROBABILITY

low-utility

Manufacture low average

Nodo 7
755.177033
High - utility

low-utility
UNFAVORABLE
0.627 Nodo 3 Subcontract low average
Nodo 8
(decision) 731.661882
High - utility

low-utility

Buy low average

Nodo 9
802.349282
High - utility

VEIM Predicted value of sample information

VEIM Valor Absoluto (VEcIM - VEsIM)
VEIM 0
VEIM 795.4 - 795.4

El resultado de tomar las decisiones es igual a 0

EFICIENCIA
E= VEIM/VEIP

0%
There is no efficiency in the information of the sample, it is considered that it should not be invested in the study, in order t
Task 2 - Solving of decision models under risks

ercise 1. DECISION TREES, EVPI and EVMI

must decide if it manufactures a new product in its main plant, or if on the contrary
of the product. The table shows projected profits, in millions of dollars.

ormation (EVPI) theory, the Expected Value of Sample Information (EVMI) and Decision

nfavorable (U) condition. The relevant conditional probabilities are:

Pagos P*P VEsIP

721 310.03 735 316.05

919 340.03
822 304.14 749.77 817 163.4

678 135.6

735 316.05

718 265.66 732.91

 732.91

756 151.2

679 291.97

919 340.03 795.4 795.4

817 163.4

s to buy with a value expected of 795,4 millions dollars

mation (VEIP) Or demand of the textile company would be 24,8 million dollars.

BABILITIES P(FՈAi) se
do en cuenta la regla de
LATER PROBABILITIES P(Ai/F)
cacion para eventos
ependientes

Se aplica regla de Bayes a cada estado

oria igual a la P(F) de la naturaleza

0.172 0.46112600536193
0.111 0.297587131367292
0.09 0.241286863270777
0.373 1

BABILITIES P(UՈAi) se
do en cuenta la regla de
LATER PROBABILITIES P(Ai/U)
cacion para eventos
ependientes

Se aplica regla de Bayes a cada estado

oria igual a la P(F) de la naturaleza

0.258 0.411483253588517
0.259 0.413078149920255
0.11 0.175438596491228
0.627 1
PROBABILITY DEMAND P*D

0.46112601 721 332.47185

0.29758713 822 244.616622 740.680965

0.24128686 678 163.592493

0.46112601 735 338.927614

0.29758713 718 213.66756 735.008043

0.24128686 756 182.412869

0.46112601 679 313.104558

Si el estudio es favarable la decisión es compr
0.29758713 919 273.482574 783.718499 783.718499 292.327

0.24128686 817 197.131367

PROBABILITY DEMAND P*D

0.41148325 721 296.679426

0.41307815 822 339.550239 755.177033

0.1754386 678 118.947368 VEcIM=

0.41148325 735 302.440191

0.41307815 718 296.590112 731.661882

0.1754386 756 132.631579

0.41148325 679 279.397129

0.41307815 919 379.61882 802.349282 802.349282 503.073

Si el estudio es desfavorable la decisión es co
0.1754386 817 143.333333
e invested in the study, in order to invest it must be above 50%.
VEcIP

819.48
s favarable la decisión es comprar.

795.4 Es el valor esperado para el nodo 1.

s desfavorable la decisión es comprar

 Unit 1 - Task 2 - Solving of decision m

Exercise 2. DECISION TREES, EV

Fabcom, a company that manufactures electronic components for the introduction inits product catalog, must dec
subcontract it with company supervision or if it buys it from an external supplier. The profits depend on the demand
dollars.

Table 2. Decision process for the commercialization of tha product

States of nature
Demand Low average -
Decision alternative
low-utility utility
demand
Manufacture 514 572
Subcontract 515 565
Buy 533 512
Lease 540 562
Leasing 541 526
Probabilities Ʃ = 1 0.31 0.29

According to the corresponding information in Table 2 and the Predicted Value of Perfect Information (EVPI) theory, the Expe
respond:

a. Use EVPI to determine if the company should try to get a better estimate of the demand.

b. A test market study of potential product demand is expected to report a favorable (F) or unfavorable (U) condition. The rele

P(F/low)= 0.3 P(D/low) = 0.7

P(F/low P(D/ low
average)= 0.4 average) = 0.6
P(F/high P(D/ high
medium) 0.43 medium) 0.57
P(F/high) = 0.41 P(D/high) = 0.59

c. What is the expected value of market research information?

d. What is the efficiency of the information?

Probability

low-utility

low average
Nodo 2
 Nodo 2
Manufacture High medium
543.38
High - utility

low-utility

low average
Nodo 3
Subcontract High medium
532.62
High - utility

low-utility

low average
Nodo 1 Nodo 4
(decision) Buy High medium
524.35
High - utility

low-utility

low average
Nodo 5
Lease High medium
560.06
High - utility

low-utility

low average
Nodo 6
Leasing High medium
525.21
High - utility

VEIP Predicted Value of perfect information

VEIP= Valor absoluto (VEcIP - VEsIP)
VEIP= 3.21

FAVORABLE
 PREVIOUS PROBABILITIES P(Ai) es la
probabilidad previa del estado de la
naturaleza CONDITIONAL PROBABILITIES
STATE OF NATURE (Ai) P(F/Ai) Es la probabilidad de que
sea favorable dado el estado de
la naturaleza (Ai)

Sumatoria igual a 1

P(F/low)= 0.31 0.3

P(F/low average)= 0.29 0.4
P(F/high medium) 0.24 0.43
P(F/high) = 0.16 0.41
1 P(F)

UNFAVORABLE
PREVIOUS PROBABILITIES P(Ai) es la
probabilidad previa del estado de la
naturaleza CONDITIONAL PROBABILITIES
STATE OF NATURE (Ai) P(U/Ai) Es la probabilidad de que
sea favorable dado el estado de
la naturaleza (Ai)
Sumatoria igual a 1

P(F/low)= 0.31 0.7

P(F/low average)= 0.29 0.6
P(F/high medium) 0.24 0.57
P(F/high) = 0.16 0.59
1 P(F)

Nodo 4
Manufacture

Nodo 5
Subcontract
 Favorable
0.3778 Nodo 2 Nodo 6
(decision) Buy

Nodo 7
Lease

Nodo 8
Leasing

Nodo 1

Nodo 9
Manufacture

Nodo 10
Subcontract
 Unfavorable
0.6222 Nodo 3 Nodo 11
(decision) Buy

Nodo 12
Lease

Nodo 13
Leasing

VEIM Predicted value of sample information

VEIM Valor Absoluto (VEcIM - VEsIM)
VEIM 0%
VEIM 560.06 - 560.06

El resultado de tomar las decisiones es igual a 0

EFICIENCIA
E= VEIM/VEIP

0% Obtendremos 0% de eficiencia.
The results obtained show that there is no efficiency in the information of the sample, therefore it is considered that it should
study it is said that it must be at least above 50%
 Unit 1 - Task 2 - Solving of decision models under risks

Exercise 2. DECISION TREES, EVPI and EVMI

on inits product catalog, must decide whether to manufacture a new product in its main plant,
The profits depend on the demand of the product. The table shows projected profits, in millions of

alization of tha product

States of nature
High medium High -
utility demand demand
utility
537 558
518 530
519 538
579 567
512 513
0.24 0.16

t Information (EVPI) theory, the Expected Value of Sample Information (EVMI) and Decision Trees,

nd.

or unfavorable (U) condition. The relevant conditional probabilities are:

Probability Pagos P*P VEsIP

0.31 514 159.34 541 167.71

572 165.88
0.29 572 165.88 579 138.96
543.38 567 90.72
 543.38
0.24 537 128.88

0.16 558 89.28

0.31 515 159.65

0.29 565 163.85

532.62
0.24 518 124.32

0.16 530 84.8

0.31 533 165.23

0.29 512 148.48

524.35
0.24 519 124.56

0.16 538 86.08

0.31 540 167.4

0.29 562 162.98

560.06 560.06
0.24 579 138.96

0.16 567 90.72

0.31 541 167.71

0.29 526 152.54

525.21
0.24 512 122.88

0.16 513 82.08

ORABLE
 JOINT PROBABILITIES P(FՈAi) se calcula
teniendo en cuenta la regla de la
LATER PROBABILITIES P(Ai/F)
multiplicacion para eventos
dependientes

Se aplica regla de Bayes a cada estado

Sumatoria igual a la P(F) de la naturaleza

0.093 0.246161990471149
0.116 0.3070407623081
0.1032 0.273160402329275
0.0656 0.173636844891477
0.3778

VORABLE

JOINT PROBABILITIES P(UՈAi) se calcula

teniendo en cuenta la regla de la
LATER PROBABILITIES P(Ai/U)
multiplicacion para eventos
dependientes

Se aplica regla de Bayes a cada estado

Sumatoria igual a la P(F) de la naturaleza
0.217 0.348762455801993
0.174 0.279652844744455
0.1368 0.219864995178399
0.0944 0.151719704275153
0.6222

Probability Demand P*P

low-utility 0.24616199 514 126.527263

low average 0.30704076 572 175.627316

545.731075
High medium 0.2731604 537 146.687136

High - utility 0.17363684 558 96.8893594

low-utility 0.24616199 515 126.773425

low average 0.30704076 565 173.478031

533.776072
High medium 0.2731604 518 141.497088

High - utility 0.17363684 530 92.0275278

low-utility 0.24616199 533 131.204341

low average 0.30704076 512 157.20487

523.596083
High medium 0.2731604 519 141.770249

High - utility 0.17363684 538 93.4166226

low-utility 0.24616199 540 132.927475

low average 0.30704076 562 172.556908 Si el estudio es favarable la decisión es alqu

562.096347 562.096347 212.36
High medium 0.2731604 579 158.159873

High - utility 0.17363684 567 98.4520911

low-utility 0.24616199 541 133.173637

low average 0.30704076 526 161.503441

523.610905
High medium 0.2731604 512 139.858126

High - utility 0.17363684 513 89.0757014

Probability Demand P*D

low-utility 0.34876246 514 179.263902 VEcIM=

low average 0.27965284 572 159.961427

541.952427
High medium 0.219865 537 118.067502

High - utility 0.1517197 558 84.659595

low-utility 0.34876246 515 179.612665

low average 0.27965284 565 158.003857

531.918033
High medium 0.219865 518 113.890068

High - utility 0.1517197 530 80.4114433

low-utility 0.34876246 533 185.890389

524.807779
 low average 0.27965284 512 143.182257
524.807779
High medium 0.219865 519 114.109932

High - utility 0.1517197 538 81.6252009

low-utility 0.34876246 540 188.331726

low average 0.27965284 562 157.164899

558.823529 558.823529 347.7
High medium 0.219865 579 127.301832 Si el estudio es desfavorable la decisión es a

High - utility 0.1517197 567 86.0250723

low-utility 0.34876246 541 188.680489

low average 0.27965284 526 147.097396

526.180971
High medium 0.219865 512 112.570878

High - utility 0.1517197 513 77.8322083

erefore it is considered that it should not be invested in the market study. to invest in a
VEcIP

563.27
s favarable la decisión es alquilar.

560.06 Es el valor esperado para el nodo 1.

s desfavorable la decisión es alquilar.
 Unit 1 - Task 2 - Solving of decision models und

Exercise 3. DECISION TREES, EVPI and EVM

Teratextyl, a textile company that has a productive experience in the foreign market of 30 years, must decide if it ma
or if on
the contrary the purchase from an external supplier. The profits depend on the demand of the product. The table show
Table 3. Decision process for the commercialization of tha product
States of nature
Demand Low average High High -
Decision alternative low-utility - medium demand
utility utility utility
demand demand

Manufacture 749 717 757 731

Subcontract 748 810 803 768
Buy 766 795 713 823
Lease 747 738 771 759
Leasing 727 814 739 721
Probabilities Ʃ = 1 0.22 0.35 0.17 0.26

According to the corresponding information in Table 3 and the Predicted Value of Perfect Information (EVPI) theory, the Expe
and Decision Trees, respond:
a. Use EVPI to determine if the company should try to get a better estimate of the demand.

b. A test market study of potential product demand is expected to report a favorable (F) or unfavorable (U) condition. The rele

P(F/low)= 0.25 P(D/low) = 0.75

P(F/low P(D/ low
average)= 0.5 average) = 0.5
P(F/high P(D/ high
medium) 0.3 medium) 0.7
P(F/high) = 0.35 P(D/high) = 0.65

c. What is the expected value of market research information?

d. What is the efficiency of the information?

Probability Pagos

low-utility 0.22 749

low average 0.35 717

Nodo 2
Manufacture High medium 0.17 757
 Nodo 2

734.48
High - utility 0.26 731

low-utility 0.22 748

low average 0.35 810

Nodo 3
Subcontract High medium 0.17 803
784.25
High - utility 0.26 768

low-utility 0.22 766

low average 0.35 795

Nodo 1 Nodo 4
(decision) Buy High medium 0.17 713
781.96
High - utility 0.26 823

low-utility 0.22 747

low average 0.35 738

Nodo 5
Lease High medium 0.17 771
751.05
High - utility 0.26 759

low-utility 0.22 727

low average 0.35 814

Nodo 6
Leasing High medium 0.17 739
757.93
High - utility 0.26 721

VEIP Predicted Value of perfect information

VEIP= Valor absolut(VEcIP - VEsIP)
VEIP= 19.66

FAVORABLE
 PREVIOUS
PROBABILITIES P(Ai) es la CONDITIONAL
PROBABILITIES P(F/Ai) Es JOINT PROBABILITIES P(FՈAi) se
probabilidad previa del calcula teniendo en cuenta la regla de
STATE OF NATURE (Ai) estado de la naturaleza la probabilidad de que
sea favorable dado el la multiplicacion para eventos
estado de la naturaleza dependientes
(Ai)

Sumatoria igual a 1 Sumatoria igual a la P(F)

P(F/low)= 0.22 0.25 0.055

P(F/low average)= 0.35 0.5 0.175
P(F/high medium) 0.17 0.3 0.051
P(F/high) = 0.26 0.35 0.091
1 P(F) 0.372

UNFAVORABLE
PREVIOUS CONDITIONAL
PROBABILITIES P(Ai) es la PROBABILITIES P(U/Ai) JOINT PROBABILITIES P(UՈAi) se
Es la probabilidad de que calcula teniendo en cuenta la regla de
STATE OF NATURE (Ai) probabilidad previa del sea favorable dado el
estado de la naturaleza la multiplicacion para eventos
estado de la naturaleza dependientes
(Ai)
Sumatoria igual a 1 Sumatoria igual a la P(F)

P(F/low)= 0.22 0.75 0.165

P(F/low average)= 0.35 0.5 0.175
P(F/high medium) 0.17 0.7 0.119
P(F/high) = 0.26 0.65 0.169
1 P(F) 0.628

Probability

low-utility

low average
Nodo 4
Manufacture High medium

High - utility

low-utility

low average
Nodo 5
Subcontract High medium

High - utility
 low-utility

Favorable low average

0.372 Nodo 2 Nodo 6
(decision) Buy High medium

High - utility

low-utility

low average
Nodo 7
Lease High medium

High - utility

low-utility

low average
Nodo 8
Leasing High medium

High - utility

Probability
Nodo 1

low-utility

low average
Nodo 9
Manufacture High medium

High - utility

low-utility

low average
Nodo 10
Subcontract High medium

High - utility

low-utility
 Unfavorable low average
0.628 Nodo 3 Nodo 11
(decision) Buy High medium

High - utility

low-utility

low average
Nodo 12
Lease High medium

High - utility

low-utility

low average
Nodo 13
Leasing High medium

High - utility

VEIM Predicted value of sample information

VEIM Valor Absoluto (VEcIM - VEsIM)
VEIM 0
VEIM 784.25 - 784.25

El resultado de tomar las decisiones es de 3,2912 si la compañía subcontrata

EFICIENCIA
E= VEIM/VEIP

0% Obtendremos 0% de eficiencia.
The results obtained show that there is no efficiency in the information of the sample, therefore it is considered that it should
study. to invest in a study it is said that it must be at least above 50%
 Solving of decision models under risks

DECISION TREES, EVPI and EVMI

30 years, must decide if it manufactures a new product in its main plant,

of the product. The table shows projected profits, in millions of dollars.

mation (EVPI) theory, the Expected Value of Sample Information (EVMI)

avorable (U) condition. The relevant conditional probabilities are:

P*P VEsIP VEcIP

164.78 766 168.52

814 284.9
803.91
250.95 803 136.51
734.48 823 213.98
128.69
 734.48

190.06

164.56

283.5
784.25 784.25
136.51

199.68

168.52

278.25
781.96
121.21

213.98

164.34

258.3
751.05
131.07

197.34

159.94

284.9
757.93
125.63

187.46
BABILITIES P(FՈAi) se
do en cuenta la regla de LATER PROBABILITIES P(Ai/F)
cacion para eventos
ependientes

Se aplica regla de Bayes a cada estado

oria igual a la P(F) de la naturaleza
0.055 0.147849462365591
0.175 0.470430107526882
0.051 0.137096774193548
0.091 0.244623655913978
0.372

BABILITIES P(UՈAi) se
do en cuenta la regla de
LATER PROBABILITIES P(Ai/U)
cacion para eventos
ependientes
Se aplica regla de Bayes a cada estado
oria igual a la P(F) de la naturaleza
0.165 0.262738853503185
0.175 0.278662420382166
0.119 0.189490445859873
0.169 0.269108280254777
0.628

Probability Demand P*P

0.14784946 749 110.739247

0.47043011 717 337.298387

730.639785
0.13709677 757 103.782258

0.24462366 731 178.819892

0.14784946 748 110.591398

0.47043011 810 381.048387 Si el estudio es favarable la decisión es subcontract

789.599462 789.599462 293.731
0.13709677 803 110.08871

0.24462366 768 187.870968

 0.14784946 766 113.252688

0.47043011 795 373.991935

786.319892
0.13709677 713 97.75

0.24462366 823 201.325269

0.14784946 747 110.443548

0.47043011 738 347.177419

748.991935
0.13709677 771 105.701613

0.24462366 759 185.669355

0.14784946 727 107.486559

0.47043011 814 382.930108

768.104839
0.13709677 739 101.314516

0.24462366 721 176.373656

Probability Demand P*D

0.26273885 749 196.791401 VEcIM= 784.25 Es el valor esperado para el nodo

0.27866242 717 199.800955

736.754777
0.18949045 757 143.444268

0.26910828 731 196.718153

0.26273885 748 196.528662

0.27866242 810 225.716561

781.08121 781.08121 490.519
0.18949045 803 152.160828 Si el estudio es desfavorable la decisión es subcontract.

0.26910828 768 206.675159

0.26273885 766 201.257962

779.377389
 0.27866242 795 221.536624
779.377389
0.18949045 713 135.106688

0.26910828 823 221.476115

0.26273885 747 196.265924

0.27866242 738 205.652866

752.269108
0.18949045 771 146.097134

0.26910828 759 204.253185

0.26273885 727 191.011146

0.27866242 814 226.83121

751.902866
0.18949045 739 140.033439

0.26910828 721 194.02707

re it is considered that it should not be invested in the market

s el valor esperado para el nodo 1.

es subcontract.
 UNIVERSIDAD NACIONAL ABIERTA Y A DISTANCIA - UNAD
Escuela de Ciencias Básicas Tecnología e Ingeniería - ECBTI

Bibliografía
 

Aven, T. (2008). Análisis de riesgos: El proceso de análisis de riesgos: Chichester, Reino

Unido: editorial de john wiley & sons (págs. 69-82). Obtenido
de http://bibliotecavirtual.unad.edu.co:2048/login?
url=http://search.ebscohost.com/login.aspx?
direct=true&db=nlebk&AN=323407&lang=es&site=ehost-live

http://home.ubalt.edu/ntsbarsh/Business-stat/otherapplets/DaRisky.htm

Pineda, R. (2019). Objeto de aprendizaje virtual Tema 1. Decisiones en un entorno de

riesgo. [Archivo de vídeo]. Obtenido de http://hdl.handle.net/10596/13263

Sharma, J. (2016). Investigación operativa: teoría y aplicaciones. (págs. 347-378), Nueva

Delhi: Laxmi Publications Pvt Ltd, v. Sexta edición. Obtenido
de http://bibliotecavirtual.unad.edu.co/login?
url=http://search.ebscohost.com/login.aspx?
direct=true&db=nlebk&AN=1950625&lang=es&site=eds-live&scope=site&ebv=EB&ppid=
pp_C

https://youtu.be/-ruRsIjPJWA