Q: 10
You use Azure Machine Learning studio to analyze an mltable data asset containing a decimal column
named column1. You need to verify that the column1 values are normally distributed.
Which statistic should you use?
Options
Discussion
C . Profile in Azure ML Studio gives you distribution charts and extra stats (like skewness) so you can actually see if it's normal, not just the average. Pretty sure that's what you'd use for this kind of check but open to hearing other takes.
Isn't "Profile" (option C) what actually gives you the distribution overview in Azure ML, not just a single stat like mean?
C . If the question narrowed it to a specific stat and not a feature, I'd switch to D, but with 'verify normal distribution' Profile is the only one showing you that directly. Someone disagree?
Option C had something like this in a mock. Profile is what you use for checking distribution.
Not D here. Mean just gives a single value, doesn't show distribution shape. Profile (C) actually lets you check for normality.
I think C for this. ML Flow tracking URI can capture a lot if you set up logging right, agree?
Option C
I'd go with ML Flow tracking URI here. It can track inputs and model outputs if the scripts log properly, so seems like a way to monitor predictions without messing much with configuration. Not entirely sure if that's the cleanest option with Azure endpoints though, maybe missing some telemetry details? Disagree if I'm off.
I don’t think it’s D. C is right since Profile actually lets you see the distribution, not just the average value.
B tbh. Not totally sure since profile gives more, but type might matter depending on context.
C imo. The Profile feature in Azure ML gives a histogram and key stats like skew and kurtosis you need to check normality, way more useful than just mean or max. Pretty sure this lines up with what the official labs recommend.
Be respectful. No spam.
