Font size
WorksheetsLesson 6 - Managed Services Machine Learning
Total questions: 10
Worksheet time: 10mins
The key challenge (s) to successfully training a machine learning Model :
Selecting features
Choosing the right algorithm
Tuning the hyper-parameter
Evaluating the metric
All of the above
Automated ML test multiple algorithm and hyper-parameter in sequence and return the best one :
True
False
To create an Automated ML run :
We need to select the algorithm first
We need to select the dataset first
We need to select the compute target first
A compute instance is a
Managed service
Not a managed service
semi managed service
none of the above
what are the things compute targets is used for
run training scripts
service deployment
Both a and b
None of the above
Select the steps involved in the end-to-end process of Machine Learning Pipelines :
Model building & training
Data preparation
Data ingestion
Model deployment
choose the correct option for the given statements :
S1 -> When DevOps applied to machine learning, we refer to the automation of machine learning pipelines as MLOps.
S2 -> DevOps is the part of MLOps.
Both True
only S1 Ture
Both False
only S2 Ture
"Calling the web service that host the model to quickly score against a relatively small set of data" is :
Real-time Inferencing
Batch Inferencing
Both
None
Select the correct option for creating real-time inferencing environments :
Create manually from Azure ML studio user interface
Create programmatically using code
Both
None
Steps you must follow while deploying a Machine Learning model used for real-time scoring, or batch inferencing :
Save and retrieve the model file in any format
Create training script
Create a real-time scoring web service
Create a schema file that describes the web service input
