MENU
Which of the following is a simple statistic to monitor for categorical feature drift?
Suggested Answer: C
Get Premium Databricks Machine Learning Professional Questions
as Interactive Web-Based Practice Test or PDF
Real Exam Format
Detailed Explanations
Instant Access
Get Your Practice Test
33 comments · newest first
KE
Kendra
10 months ago
Percentage of missing values is also important, but not the only one.
DO
Dominga
11 months ago
Totally agree with C, it gives a fuller picture!
TO
Tommy
11 months ago
Wait, can you really just use the mode for feature drift? Seems too simple.
MA
Margo
11 months ago
I think option C covers more ground, though.
CE
Celestina
11 months ago
The mode is definitely a key statistic for categorical data!
AL
Alesia
12 months ago
I vaguely remember that the percentage of missing values can indicate drift, but I'm not confident if it's the best choice here.
WI
Winfred
12 months ago
I feel like the answer might be C since it includes multiple aspects of monitoring, but I can't recall if all of them are necessary.
BE
Bettina
12 months ago
I think we practiced a question about feature drift, and the mode was definitely mentioned as important.
JA
Janine
1 year ago
I remember we discussed the mode as a key statistic for categorical data, but I'm not sure if it's the only one we should focus on.
ST
Stephaine
1 year ago
I'd go with the percentage of missing values. That's a simple but important metric to keep an eye on, as changes there could indicate issues with data collection or processing.
TH
Thad
1 year ago
The mode, number of unique values, and percentage of missing values seem like a good set of statistics to track. That would give you a pretty comprehensive view of any changes in the categorical feature over time.
TR
Trinidad
1 year ago
I'm a bit unsure about this one. I know the mode is a measure of central tendency, but I'm not sure if it's the best for monitoring drift. Maybe the number of unique values or percentage of missing values would be better?
MA
Mammie
1 year ago
I think the mode is a good simple statistic to monitor for categorical feature drift. It gives a sense of the most common value, which could change over time.
VE
Verdell
1 year ago
I'm leaning towards the number of unique values as the best option here. Tracking how the diversity of the feature changes over time could be a useful way to detect drift.
FR
Frederica
1 year ago
This is a tricky one. There are a few options that seem relevant, but I'm not sure which one is the best fit. I'll need to think through the requirements step-by-step to make sure I select the right answer.
DA
Dawne
1 year ago
I'm a bit unsure about the third one, 'not' not in 'in'. I'll need to double-check the logic on that one.
TA
Tammara
1 year ago
Hmm, I wonder if the mode is the most suitable statistic for monitoring categorical feature drift. Maybe we need to consider the number of unique values and the percentage of missing values too. Just a thought.
LA
Lacey
1 year ago
It's good to have multiple statistics to monitor for a more comprehensive analysis.
MI
Mi
1 year ago
I agree, the mode alone may not be enough to monitor categorical feature drift.
VI
Viki
1 year ago
I think considering the number of unique values and percentage of missing values is important too.
LO
Lorean
1 year ago
The mode? Really? I thought we were supposed to monitor the number of unique values and missing values. This is a no-brainer, folks. Option C is the way to go.
JA
Jackie
1 year ago
Let's make sure to keep an eye on the number of unique values and missing values to catch any drift early on.
KA
Katy
1 year ago
Yeah, option C seems like the most comprehensive choice to monitor categorical feature drift.
BR
Brittani
1 year ago
I think the mode can also be useful in some cases, but overall, option C covers all the important statistics to monitor.
TA
Tamekia
1 year ago
I agree, monitoring the number of unique values and missing values is crucial for detecting drift.
FL
Fletcher
1 year ago
Option C covers all the bases. Monitoring those three statistics should give a good indication of any changes in the categorical feature over time.
BO
Bobbye
1 year ago
I'm not sure if the mode alone is enough. Checking the number of unique values and percentage of missing values as well seems like a more comprehensive approach.
KR
Kristeen
1 year ago
The mode seems like a good starting point to monitor for categorical feature drift. I wonder if the number of unique values and percentage of missing values could also provide useful insights.
NY
Nydia
1 year ago
Yes, keeping an eye on the number of unique values and percentage of missing values can also help us detect any drift in the categorical feature.
EN
Ena
1 year ago
I agree, monitoring the mode can give us a good idea of any changes in the most frequent category.
NI
Nieves
2 years ago
But monitoring the percentage of missing values is also important to detect drift in categorical features.
MY
My
2 years ago
I disagree, I believe the answer is E) Number of unique values.
NI
Nieves
2 years ago
I think the answer is C) Mode, number of unique values, and percentage of missing values.