Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit

BUSINESS ANALYTICS

METHODS OF DECISION ANALYSIS (QBUS3330)

INDIVIDUAL ASSIGNMENT  SEMESTER 2 2022

Question 1 (75 marks):

Xenon Chemicals has developed an efficient product to treat mold in houses and apartments. Before launching the new product, they face the following decisions:

•   They have the option to improve the product by either purchasing a new technology or by attempting to develop the new technology.

•   Purchasing the new technology entails purchasing a small R&D company at a   cost of 60 million dollars. In this case, they are certain to improve their product.

•   Xenon may choose to invest 35 million dollars in attempting to develop the   technology in-house. The head of R&D at Xenon estimates the probability of success at 0.65.

•   If R&D is successful, then Xenon can sell an improved product. If it is             unsuccessful, they can still sell the existing product (which although inferior is still quite efficient).

•   Clearly, Xenon may choose to neither purchase the technology nor attempt to develop it. If this is the case, they will market the existing product.

In order to simplify the analysis, the marketing manager of Xenon estimates that sales may be: High, Moderate or Low. She believes that the following revenues are likely   given the type of product Xenon decides to sell:

Existing Product (Existing Technology)

Revenue (Millions of $)

140

70

20

Probability

0.4

0.3

0.3

Upgraded Product (New Technology)

Revenue (Millions of $)

180

120

20

Probability

0.5

0.4

0.1

Before deciding on their strategy, Xenon have been approached by an international        chemical company, Argon. Argon is offering to partner with Xenon and has put forward the following offer:

•   If Xenon decides to sell the existing product (without attempting to develop a new technology), then Argon will pay Xenon 5 million dollars upfront and a further     $110M, $70M and $40M in the event of high, moderate or low sales, respectively.

•   If Xenon decides (and is able) to sell the new product (new technology), then       Argon will pay Xenon 30 million dollars upfront and a further $150M, $90M and $50M in the event of high, moderate or low sales, respectively.

Assume that Argon’s involvement does not affect the probabilities of high, moderate, or low sales. Furthermore, you can assume that Xenon can decide whether or not to partner with Argon after having resolved the decision whether or not to upgrade their                 technology.

Section 1: Describe the decision situation using a decision tree. (We strongly recommend you use Precision Tree or a similar software package).

Section 2: What is optimal policy ifXENON’s sole objective is to maximise profit from marketing the product? What is the EMV of profit under this strategy.

Section 3: Analyse the problem using the Stochastic Dominance criterion. Are your        results consistent with Section 2? Only consider 3 courses of action: use existing              technology, attempt to develop a new technology and purchase a new technology             (assume that you decide whether to partner with ARGON optimally for each strategy).    Section 4: What is the optimal policy if XENON wants to maximise the probability of a profit of at least 60M dollars? What is XENON’s profit under this policy?

Section 5: Assume that negotiations with ARGON are unsuccessful, and XENON            decides that it will not partner with ARGON before making its decision whether or not to purchase new technology, attempt developing it or use the existing technology. What is   its optimal course of action? What is the expected profit under the policy?

Question 2 (25 marks):

A large semi-conductor producer finds that, based on historical data, 10% of microchips it produces are faulty and have to be replaced. The producer can decide not to address the problem, resulting in the need to replace 10% of microchips it sells.

Alternatively, the producer has the option to add a second stage of production for some, or all, microchips. If a micro-chip is faulty then there is a 0.8 probability it will be fixed in the second stage of production. There is also a 0.2 probability it will remain faulty. If a microchip is in working condition (not faulty) then there is a 0.1 probability that the second stage of production will render it faulty. There is a 0.9 probability it will remain a working microchip.

A newly developed QA test can help determine whether a microchip is faulty or in working condition. A positive test results suggests that it is in working condition, while a negative test result implies that it is believed to be faulty. Based on extensive laboratory testing, it was found that 86.5% of tests resulted in positive results. The remainder of tests resulted in negative results as there are no inconclusive outcomes. Based on the same testing, it was determined that given a positive test there is a 0.95 probability that the chip is in working condition, and given a negative test there is a 0.9 probability it is faulty.

Section 1: What are the alternatives the semi-conductor producer should consider if the goal is to minimize the likelihood of faulty microchips it sells? Describe the decisions they face using a decision tree.

Section 2: What is the preferred course of action that minimizes the number (or         proportion) of faulty microchips sold? What percentage of microchips will need to be replaced?