This preview shows page 1. Sign up to view the full content.
Lab for Statistical Decision Theory
Data Explanation:
Same as the data used in Practicum 3
Problem 1
For the target TAR1,
(a) Repeat Practicum 2 with exactly the same options used in the practicum.
(b) Write down the profit equation for each decision where p
2
is the probability for
"Tar1=2" and p
1
is the probability for "Tar1=1'?
(Note: probability for TAR1=0=1
p
1
p
2
)
(c) Suppose the company can mail all customers without coupon.
Create a variable
representing the profit of this action. (Note: The revenue for one ordmail is $20 and it
cost $0.5 per mail)
(d) Suppose the company can mail all customers with coupon.
Create a variable
representing the profit for this action. (Note: The coupon worth $2 for each customer
who responses and the customer suppose to buy 10% more if there is a coupon.)
(e) Create a variable representing perfect prediction with coupon.
(f) Use variable "CP_tar1_" representing the profit based on the model profit matrix or
use ordmail to compute the profit representing the model profit matrix.
(g) Add another Data Source (use data set CAMPSCR), one Score Node, and one SAS
This is the end of the preview. Sign up
to
access the rest of the document.
This note was uploaded on 10/04/2011 for the course STA 6714 taught by Professor Staff during the Spring '11 term at University of Central Florida.
 Spring '11
 Staff

Click to edit the document details