Please help me to solve these problems.

9. Transform the following salaries to numerical scales with $20,000 equal to 0,

$220,000 equal to 1.0, and everything else linearly related between 0 and 1.

Client Salary

Fred $120,000

Herman $200,000

George $50,000

Frieda $65,000

Hermione $35,000

11. Given the following data, alternative data mining models have been appliedon a test set of data.

What appears to be the best model?

Subject Actual Regression Clustering NeuralNet Rule-Based

Fred good good good good good

Herman bad good bad good bad

George good good bad bad bad

Frieda good good good good bad

Hermione bad good bad bad good

12. Do the coincidence matrix for each of the four models in problem11.

13. In problem 12, the regression and neural net models had the same number oferrors.

Why might the results of one model be considered better than theresults of the other?

14. In problem 11, given relative error costs of the model calling an actually goodoutcome bad to be

$100 and calling an actually bad outcome good to be $500,

what is the total cost of errors for each of the four models?