Book we are using is called: Management Science: The Art of Modeling with Spreadsheets

1. Download the following file concerning heart attacks in South Africa. I do believe the database is clean with no problems. Let me know if something happens.

a. Read the data dictionary

b. Use an "If" statement to create a 1/0 variable pertaining to family history. Make sure the column is adjacent to the existing data. Make sure you place a label in the top row.

c. Produce descriptive statistics for each variable. What do you conclude?

d. Produce a correlation matrix. What do you conclude?

e. Prepare a scatterplot comparing adiposity and obesity. What does it tell you?

f. Prepare a scatterplot matrix for all variables. What does it tell you?

g. Prepare histograms for each variable that is not a 1/0 variable. What do they tell you?

h. Prepare a boxplot against the variable "chd." What does this tell you?

2. Partition the data REMEMBER THIS!!!!

A. Prepare classification analyses for each of the following

i. K-NN using 10 nearest neighbors (remember to have it analyze 1-10
ii. Classification tree
iii. Logistics regression

B. What conclusions do you draw?

Chapters 8 through 11

3. What is the objective function in optimization telling you?

4. What is a constraint and what is its impact that makes it important?

5. What is the difference between a binding constraint and non-binding constraint?

6. What does a shadow price tell us?

7. Complete the following problem found at this excel file Optimization Problem for Final. Follow the instructions on the spreadsheet.

8. Look over the example on page 258 (Bonner Electronics). Explain what optimization achieves in this problem and the relationship between supply and demand in the model.

Bonner Electrics worksheet can be found on: www.wiley.com/college/powell.

9. Why are integer and binary optimization models useful?

Chapter 13

10. What is the difference between the benchmark methods (see page 337-340) and using probabilities?

11. Complete the problem in this file. Decison Matrix for Final. This is problem #3 in chapter13.

I have computed the payoff amounts for each cell. Answer a., b., c., d., and e.

Chap13 #3 problem:

Jr. Davidson recently started a practice in a landscape design and is considering the purchase of an automated drafting system. Jr. can purchase a system with three possible drafting capacities. The payoffs for having any of these systems depend on the demand for drafting services over the next few years. The cost for each system are shown as follows along with Jr.'s assessment of the probabilities that demand will match the capacity of each one:

Total cost
probability
Small system
$10,000
0.4
Medium system
$14,000
0.3
Large system
$20,000
0.3

Working capacity, each system would generate net cash flow at a yearly rate of 50 % of its total cost. If a system is chosen that is smaller than demand, it would work at capacity. If a system is chosen that is larger than demand, revenue from the system would be limited by demand. For convenience, JR has initially decided to count cash flow for three years, without discounting. For example, if JR chooses the medium system and demand is Small, then the profit is calculated as follow: profit=3(0.5*10,000)-14,000=$1,000

A. What is the best decision under the maximax criterion?

B. What is the best decision under the maximin criterion?

C. What is the best decision under the minimax regret criterion?

D. What is the best decision under the expected payoff criterion?

E. Reviewing the analysis, JR decides that the assumption of a 3-years horizon is too restrictive. Instead, it makes more sense to threat the horizon as uncertain, with the following probability distribution:

Two years of cash flow has 0.4 probability

Three years of cash flow has 0.4 probability

Four years of cash flow has 0.2 probability

Now, what is the best decision under the expected payoff criterion?