MENU
A data scientist has produced three new models for a single machine learning problem. In the past, the solution used just one model. All four models have nearly the same prediction latency, but a machine learning engineer suggests that the new solution will be less time efficient during inference.
In which situation will the machine learning engineer be correct?
Suggested Answer: D
Get Premium Databricks Machine Learning Associate Questions
as Interactive Web-Based Practice Test or PDF
Real Exam Format
Detailed Explanations
Instant Access
Get Your Practice Test
If the new solution requires that each of the three models computes a prediction for every record, the time efficiency during inference will be reduced. This is because the inference process now involves running multiple models instead of a single model, thereby increasing the overall computation time for each record.
In scenarios where inference must be done by multiple models for each record, the latency accumulates, making the process less time efficient compared to using a single model.
34 comments · newest first
MI
Mirta
10 months ago
E seems off, size doesn't always mean slower performance, right?
ME
Meghan
11 months ago
B could be a factor too, but A is definitely a big issue.
JO
Jospeh
11 months ago
Wait, can’t the models still be efficient even with if-else?
NA
Natalie
11 months ago
Totally agree, A makes the most sense here.
HO
Hollis
11 months ago
I think it's A. If-else logic can slow things down.
KE
Kerrie
12 months ago
I’m a bit confused about the sizes of the models, but I think option E might be relevant if the new models are larger than the original one.
RE
Reita
12 months ago
I practiced a question similar to this, and I think option D could be correct if each model has to compute predictions for every record.
FI
Fidelia
12 months ago
I'm not entirely sure, but I feel like the average latency of the new models could be a factor, maybe option B?
LE
Leonora
1 year ago
I remember discussing how if-else logic can slow down inference, so I think option A makes sense.
EM
Emerson
1 year ago
This seems straightforward. The engineer's concern about time efficiency during inference is likely to be correct when the new solution requires some kind of additional logic or processing to determine which model to use. I'll go with option A.
DE
Deonna
1 year ago
Alright, I think I've got it. The key is to identify the situation where the use of multiple models would actually make the inference process less efficient. Time to carefully consider each option.
IZ
Izetta
1 year ago
I'm a bit confused by the wording here. What exactly does "time efficient during inference" mean? I'll have to make sure I understand that before I can decide on the right answer.
JA
Jacqueline
1 year ago
Okay, let me see. The key seems to be understanding how the new solution will impact inference time efficiency. I'll need to consider the different factors mentioned.
HE
Herminia
1 year ago
Hmm, this seems like a tricky one. I'll need to think carefully about the implications of using multiple models versus a single model.
VI
Virgina
2 years ago
You know, I think the correct answer is when the new solution requires each model to compute a prediction for every record. That's got to be less efficient.
FL
Florencia
2 years ago
Haha, I bet the engineer just wants to keep using the old model. Gotta love that resistance to change!
TI
Titus
2 years ago
D) When the new solution requires that each model computes a prediction for every record
SA
Santos
2 years ago
C) When the new solution requires the use of fewer feature variables than the original model
GR
Gregg
2 years ago
B) When the new solution's models have an average latency that is larger than the size of the original model
WA
Walker
2 years ago
A) When the new solution requires if-else logic determining which model to use to compute each prediction
RO
Ronny
2 years ago
Wait, what if the new models are just bigger in size? That could definitely slow things down during inference.
HO
Hollis
2 years ago
B) When the new solution's models have an average size that is larger than the size of the original model
AL
Alyssa
2 years ago
A) When the new solution requires if-else logic determining which model to use to compute each prediction
GE
German
2 years ago
But what if the new solution's models have an average size larger than the original model? Wouldn't that also make it less time efficient?
TR
Tracey
2 years ago
Nah, I don't think that's the case. If the latency is the same, the extra logic shouldn't make a difference.
AL
Aliza
2 years ago
Nah, I don't think that's the case. If the latency is the same, the extra logic shouldn't make a difference.
LO
Louisa
2 years ago
A) When the new solution requires if-else logic determining which model to use to compute each prediction
JU
Junita
2 years ago
Hmm, I think the engineer is right. If the new solution requires if-else logic, that's gonna add some overhead, right?
SA
Sage
2 years ago
D) When the new solution requires that each model computes a prediction for every record
MY
My
2 years ago
C) When the new solution requires the use of fewer feature variables than the original model
DA
Daron
2 years ago
B) When the new solution's models have an average latency that is larger than the size of the original model
LO
Louvenia
2 years ago
A) When the new solution requires if-else logic determining which model to use to compute each prediction
FL
Floyd
2 years ago
I agree with Stefany. If there's if-else logic, it will definitely slow down the inference process.
ST
Stefany
2 years ago
I think the machine learning engineer will be correct when the new solution requires if-else logic to determine which model to use.