Term
| With incremental response models two models are used. 1 calculates the propensity to respond ___. Model 2 calculates the propensity to respond to ___. Calculate the difference between the two and use ___. |
|
Definition
without a message
a specific message
the most effective |
|
|
Term
| incremental response models estimate the increments in response based on the ___ |
|
Definition
|
|
Term
| Define these responses to incremental modeling. Sure things persuadables lost causes sleeping dogs |
|
Definition
| 1. will respond to the offer regardless if they got the message 2. going to respond, but only if they get the offer 3. who aren't going to respond regardless of the offer 4. who are going to do something nasty like quit. |
|
|
Term
| ___ is the ability of a model to work, not only on the model set, but also on unseen data |
|
Definition
|
|
Term
| what are the enemies of stability |
|
Definition
getting things wrong
overfitting
sample bias
future may not be like the past |
|
|
Term
| what are profiling models based on |
|
Definition
demographic variables which include:
geographic location
gender
age |
|
|
Term
| what are the 11 steps to the directed data mining methodology |
|
Definition
1 translate the business problem into a data mining problem
2. select appropriate data
3. get to know the data
4. create a model set
5. fix problems with the data
6. transform data to bring information to the surface
7. build models
8. assess models
9 deploy models
10. assess results
11. begin again |
|
|
Term
| ___ is how well the models do at identifying important outcomes |
|
Definition
|
|
Term
| ___ is how well the models do at identifying other outcomes |
|
Definition
|
|
Term
| The particular measure for a given system was the area under the ROC curve, which is called ___ |
|
Definition
|
|
