Marketing-AW108 Online Services
Business Intelligence
Ques #1 Given the following Sales data in sales table. It consists of six products (Apples, Shampoo, Napkins, Coffee, Soap, Toothpaste) and each transaction has at the most 2 products implying there are no combination of 3 products.
Assume a threshold level of 2 and 75% confidence requirement.
Hint: while counting make sure to look at both columns. Example some purchases maybe Apples, shampoo or toothpaste, Apples
Apples must be counted in both columns
Sales data
Ctrl+ click
- Perform apriori analysis and provide the confidence that a customer will buy toothpaste given they have already bought shampoo. Is this rule acceptable?
- Provide the confidence that a customer will buy Toothpaste given they have already bought soap. Is the rule acceptable?
- Provide all acceptable rules at 75% confidence
Ques#2. (10 points)
Given the following ownership data for Tractorsbased on income level and lot size
(similar example that we did in class but NOT the same)
Ctrl+ click
Develop a two level split decision tree using
First level split at lot size = 19
Second level split follow following rule
If lot size > 19 AND income <= 75.5 etc..
If lot size<=19 AND income <= 65.8 etc..
Do not worry about third level splits
After developing decision tree calculate
Gini index
- Before the split
- After the first split (size =19)
- After the second level splits
Q3 Given the following
Given the following Table with training Records with their actual class and the probability of them being class 1 members estimated by a classifier. Where 1 denotes a customer will buy washer/dryer and 0 implies they will not.
(12 points)
(sourceShmuelietc)
| Probability of class 1 | Actual Class |
| 0.9950 | 1 |
| 0.9875 | 1 |
| 0.9844 | 1 |
| 0.9804 | 1 |
| 0.9481 | 0 |
| 0.8893 | 1 |
| 0.8476 | 1 |
| 0.7628 | 0 |
| 0.7070 | 1 |
| 0.6807 | 1 |
| 0.6563 | 1 |
| 0.6224 | 0 |
| 0.5050 | 0 |
| 0.4713 | 1 |
| 0.3371 | 1 |
| 0.2179 | 0 |
| 0.1992 | 0 |
| 0.1494 | 0 |
| 0.0478 | 1 |
| 0.0383 | 1 |
| 0.0248 | 0 |
| 0.0218 | 0 |
| 0.0161 | 0 |
| 0.0035 | 0 |
- Draw a “lift” chart for a person buying a washer and dryer (1) using the above model and provide the lift value for 10 cases o
Q4. Given the following decision tree
As noted in class split values are arbitrary in the example. In real situation an algorithm will define split
| MAYBE |
| 2 |
| 2 |
| NO |
| GMAT |
| 6 |
| 650 |
| 2 |
| 1 |
| 3.8 |
| 3 |
| 6 |
| 3 |
| 3.25 |
| 580 |
| 8 |
| GPA |
(3 points)
| YES |
| NO |
| GMAT |
| MAYBE |
| GPA |
| NO |
| GPA |
| 2 |
| 4 |
| 2.5 |
| Yes |
The designer has decided to prune the tree after three nodes GPA=3.25, GMAT=650 and
after GPA = 3.25 and GMAT=580.
Redraw the tree after pruning; you can copy this and use edit (cut/copy/paste etc..) in WORD to make changes to the tree. Or hand draw it and scan it.
( 3 points)
Product code: Marketing-AW108
Looking for best Marketing-AW108 online ,please click here