MENU
One common approach for predicting rare events in the LOGISTIC procedure is to build a model that disproportionately over-re presents those cases with an event occurring (e.g. a 50-50 event/non-event split).
What problem does this present?
Suggested Answer: B
Get Premium A00-240 Questions
as Interactive Web-Based Practice Test or PDF
Real Exam Format
Detailed Explanations
Instant Access
Get Your Practice Test
12 comments · newest first
MI
Micaela
12 months ago
Not sure about this, can we really trust those estimates?
DA
Dan
12 months ago
D makes sense too, sensitivity can get messed up.
NA
Nakisha
1 year ago
Wait, only the intercept? That seems off.
IR
Ira
1 year ago
Totally agree, bias is a big issue here!
ET
Ettie
1 year ago
A is the right answer, all estimates get skewed.
FI
Filiberto
1 year ago
I’m leaning towards the idea that sensitivity estimates could be biased too, especially if the model is skewed.
OC
Octavio
1 year ago
I feel like I read somewhere that only the non-intercept estimates could be biased, but I can't recall the details.
AN
An
1 year ago
I think we practiced a similar question where we talked about how the intercept might be affected more than other estimates.
LA
Latosha
1 year ago
I remember discussing how oversampling can lead to biased parameter estimates, but I'm not sure if it's all parameters or just some.
AL
Alison
1 year ago
Hmm, this one seems tricky. I'll need to carefully review the options and think through the possible configurations for clientless SSLVPN.
NI
Nickole
1 year ago
Hmm, this one seems tricky. I'll have to think it through carefully.
MA
Marisha
1 year ago
Okay, I think I've got this. Based on the requirements, I'd say we need to apply the Data Format Transformation, Data Model Transformation, and Event-Driven Messaging patterns in addition to the Orchestration compound pattern. The other options don't seem as directly relevant.