Phase 2 - Solve problems by applying the algorithms of the Unit 1 Sofia Martinez 26 septiembre 2019 Universidad Nacion
Views 93 Downloads 2 File size 223KB
Phase 2 - Solve problems by applying the algorithms of the Unit 1
Sofia Martinez 26 septiembre 2019
Universidad Nacional Abierta y a Distancia Escuela de Ciencias Básicas, Tecnología e Ingeniería Competencias comunicativas
Exercise 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 alternative
Manufacture Subcontract Buy Probabilities Ʃ = 1
Demand lowutility
Demand low average - utility
321 335 318 0,35
352 353 361 0,35
Demand High utility 367 381 375 0,3
According to the corresponding information in Table 1 and the Predicted Value of Perfect Information (EVPI) theory, the Expected Value of Sample Information (EVMI) 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 relevant conditional probabilities are: P(F/low) = 0,35
P(D/low) = 0,65
P(F/low average) = 0,41
P(D/ low average) = 0,59
P(F/high) = 0,32
P(D/high) = 0,68
a. What is the expected value of market research information? b. What is the efficiency of the information?
Solución a.
Use EVPI to determine if the company should try to get a better estimate of the demand.
BAJA (0,35)
FABRICAR
321
112
MEDIA NODO BAJA(0,35) 2
352
123
MEDIA ALTA (0,3)
367
110
BAJA (0,35)
335
117
MEDIA BAJA (0,35)
353
124
MEDIA ALTA (0,3)
381
114
BAJA (0,35)
318
111
MEDIA BAJA (0,35)
361
126
MEDIA ALTA (0,3)
375
113
NODO SUBCONTRATAR NODO 1 3
COMPRAR
NODO 4
345
355
350
Nodo 2 Nodo 3 Nodo 4 Nodo 1
345 355 350 355
La decisión recomendada por el método del valor esperado es fabricar con un pago de 355 millones de dolares. EVPI = EVwPI – EVwoPI EVwPI = (0.35x335) + (0,35x361)+(0,3x381) = 357,9 EVwoPI = 355 EVPI = 357,9 – 355 = 2, 9 = 3
b. A test market study of potential product demand is expected to report a favorable (F) or unfavorable (U) condition. The relevant conditional probabilities are: P(F/low) = 0,35
P(D/low) = 0,65
P(F/low average) = 0,41
P(D/ low average) = 0,59
P(F/high) = 0,32
P(D/high) = 0,68
Teorema de Bayes
P(A₁) = 0,35 , P(A₂) = 0,35
P(A₃) = 0,3
FAVORABLE Estados de la naturaleza
Probabilidades previas
Probabilidades condicionales P(F)
Probabilidades conjuntas
Probabilidades posteriores
ALTA MEDIA BAJA
0,35 0,35 0,3
0,35 0,41 0,32
0,1225 0,1435 0,096
0,33839779 0,39640884 0,26519337
P(F)
0,362
FAVORABLE Estados de la naturaleza
Probabilidades previas
Probabilidades condicionales P(F)
Probabilidades conjuntas
Probabilidades posteriores
ALTA MEDIA BAJA
0,35 0,35 0,3
0,65 0,59 0,68
0,2275 0,2065 0,204
0,356583072 0,323667712 0,319749216
P(F)
0,638
FABRICAR
FAVORABLE (O,362)
NODE 2
SUBCONTRATA R
COMPRAR
NODO 4
NOD0 5
NODO 6
NODE 1 FABRICAR
DESFAVORABL E (0,638)
NODE 3
SUBCONTRATA R
COMPRAR
NODO 7
NODO 8
NODO 9
BAJA
0,34
321
MEDIA BAJA
0,40
352
MEDIA ALTA
0,27
367
BAJA
0,34
335
MEDIA BAJA
0,4
353
MEDIAALTA
0,27
381
BAJA
0,34
318
MEDIA BAJA
0,4
361
MEDIAALTA
0,27
375
BAJA
0,36
321
MEDIA BAJA
0,32
352
MEDIAALTA
0,32
367
BAJA
0,36
335
MEDIA BAJA
0,32
353
MEDIAALTA
0,32
381
BAJA
0,36
318
MEDIA BAJA
0,32
361
MEDIAALTA
0,32
375
VE (Nodo 4) = 349,03 VE (Nodo 5) = 357,97 VE (Nodo 6) = 353,77 VE (Nodo 7) = 345,64 VE (Nodo 8) = 355,48 VE (Nodo 9) = 350 Favorable (Nodo 2) = Max ( Nodo4, Nodo5, Nodo6) = 367,97 Desfavorable (Nodo 3) = Max ( Nodo7, Nodo8, Nodo9) = 355,48
349,03
357,97
353,77
345,64
355,48
350
a.
What is the expected value of market research information?
VE (Nodo1) = (0,362*367,97) + (0,638*355,48) = 360,001 b.
What is the efficiency of the information?
Valor de la información perfecta: VEIM = VecIM – VesIM = 360,001 - 355 = 5,001 E = (VEIM/VEIP) x 100 = (5,001 / 3 ) x 100 = 166,7 %
Exercise 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.
According to the corresponding information in Table 2 and the Predicted Value of Perfect Information (EVPI) theory, the Expected Value of Sample Information (EVMI) 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 relevant conditional probabilities are:
P(F/low) = 0,15
P(D/low) = 0,85
P(F/low average) = 0,45
P(D/ low average) = 0,55
P(F/high medium) = 0,3
P(D/ high medium) = 0,7
P(F/high) = 0,5
P(D/high) = 0,5
c. What is the expected value of market research information? d. What is the efficiency of the information?