IBM SPSS MODELER COURSE 1 UNIT 1 & 2 EXAM
QUESTIONS AND ANSWERS WITH COMPLETE
SOLUTIONS GRADED A++ LATEST UPDATE
Fill in the blank. There are ___ stages in CRISP-DM.
There are 6 stages in the CRISP-DM methodology.
You work as a data scientist in a project where you need to access certain tables
in the database. You ask the database administrator (DBA) for access to the
database and he promises to do so as soon as possible.
However, a minute later the company experiences serious problems in the
operational systems, which also fall under the DBA's responsibility. It is a top
priority for the DBA to get the operational systems up and running again, which
will cost 2 days at a minimum. In the meantime you are stuck in the project,
facing a deadline.
Could such a situation occur in a project where CRISP-DM is implemented?
A. Yes
B. No
B. No. Problems such as these should be acknowledged in the Business
Understanding stage, and properly dealt with at that point.
You work as a data scientist in a project, about reducing churn (caused by
customers canceling their subscription). After two months of collecting data, data
, preparation, and analyses, you have built a model that identifies 95% of the
customers who canceled their subscription in 10% of the data.
Your recommendation to the manager is to apply the model to the current
customer database, which will identify 95% of the potential churn (that is the
expectation based on the analyses). However, the manager tells you that 100% of
the potential churn must be identified, no matter what, and therefore the model
that you have built will not be deployed.
Could such a situation happen in a project where CRISP-DM is
implemented?
A. Yes
B. No
B. No. It should be clear upfront that the manager only accepts a model that identifies
100% of the customers who cancel their subscriptions. The modeling goal should have
been addressed in the Business Understanding stage. Also, it should be discussed
whether the modeling goal set by the manager is realistic.
You work in the database marketing department of ACME, a (fictitious) company
that sells sport products via the web and mail campaigns. ACME has 30,000
customers.
You are responsible for a campaign to promote and sell the long awaited XL
Original Orange Baseball Cap. Given the costs of contacting a customer and to
prevent sending unsolicited mail, you will send an e-mail to 10,000 randomly
selected customers with an offer to buy the XL Original Orange Baseball Cap.
This e-mail campaign will be referred to as "test mailing" from this point forward.
QUESTIONS AND ANSWERS WITH COMPLETE
SOLUTIONS GRADED A++ LATEST UPDATE
Fill in the blank. There are ___ stages in CRISP-DM.
There are 6 stages in the CRISP-DM methodology.
You work as a data scientist in a project where you need to access certain tables
in the database. You ask the database administrator (DBA) for access to the
database and he promises to do so as soon as possible.
However, a minute later the company experiences serious problems in the
operational systems, which also fall under the DBA's responsibility. It is a top
priority for the DBA to get the operational systems up and running again, which
will cost 2 days at a minimum. In the meantime you are stuck in the project,
facing a deadline.
Could such a situation occur in a project where CRISP-DM is implemented?
A. Yes
B. No
B. No. Problems such as these should be acknowledged in the Business
Understanding stage, and properly dealt with at that point.
You work as a data scientist in a project, about reducing churn (caused by
customers canceling their subscription). After two months of collecting data, data
, preparation, and analyses, you have built a model that identifies 95% of the
customers who canceled their subscription in 10% of the data.
Your recommendation to the manager is to apply the model to the current
customer database, which will identify 95% of the potential churn (that is the
expectation based on the analyses). However, the manager tells you that 100% of
the potential churn must be identified, no matter what, and therefore the model
that you have built will not be deployed.
Could such a situation happen in a project where CRISP-DM is
implemented?
A. Yes
B. No
B. No. It should be clear upfront that the manager only accepts a model that identifies
100% of the customers who cancel their subscriptions. The modeling goal should have
been addressed in the Business Understanding stage. Also, it should be discussed
whether the modeling goal set by the manager is realistic.
You work in the database marketing department of ACME, a (fictitious) company
that sells sport products via the web and mail campaigns. ACME has 30,000
customers.
You are responsible for a campaign to promote and sell the long awaited XL
Original Orange Baseball Cap. Given the costs of contacting a customer and to
prevent sending unsolicited mail, you will send an e-mail to 10,000 randomly
selected customers with an offer to buy the XL Original Orange Baseball Cap.
This e-mail campaign will be referred to as "test mailing" from this point forward.
