Hello Readers! Here are some questions to help you design and implement data science solutions on Microsoft Azure.
Answer the following questions after you have read the course material. The correct answers can be found at the end.
Here’s the Quiz:
Which of the following is a web interface that allows you to manage assets in a workspace? Azure Machine Learning studiob. Azure Cognitive Servicesc. Azure Synapse Analytics
Automated machine learning is used with data from car sales to train a machine-learning model that predicts the car’s price based on its make, size, engine size, mileage, and other factors. Which task type should you choose? Classificationb. Regressionc. Time-series
Riya uses the Azure Machine Learning Python Python SDK to create code for an experiment. She needs to be able to easily retrieve the metrics from each run. To print metrics, add print statements to the experiment code. To record named metrics, use the log methods of Run class. Save the experiment data to the outputs folder.
Mahesh has a reference for a Workspace called ws.Which Code retrieves the default datastore of the workspace?a. default_ds = ws.get_default_datastore()b. default_ds = Datastore.get(ws, ‘default’)c. default_ds = ws.Datastores[0]
What ScriptRunConfig parameter causes a script to run on a cluster of compute named train-cluster.a. environment=’train-cluster’b. compute_target=’train-cluster’c. arguments=[‘-AmlCluster’, ‘train-cluster’]
Which type of object should be used to pass data between pipeline steps. Datastoreb. Datasetc. PipelineData
You are looking to implement a batch-inference pipeline that distributes scoring across multiple nodes. PythonScriptStepb. AdlaStepc. ParallelRunStep
Automated machine learning is used to find the best model with the highest AUC_weighted. task=’AUC_weighted’b. label_column_name= ‘AUC_weighted’c. primary_metric=’AUC_weighted’
A binary classification model is being trained to support college admission decisions. Each model should be evaluated using a validation dataset. The model with the highest accuracy score should be used. Remove the ethnicity feature in the training dataset. Compare the disparity in selection rates and performance metrics across ethnicities.
You have previously trained a model with a training dataset. What should you do to detect data drift in the new data since the model was trained? Create a new dataset with the new data and a timestamp columns. Then create a data drift monitor using the training dataset as a baseline, and the new dataset to target. Retrain the model by creating a new version using only the new data. Add the new data to an existing dataset and enable Application Insights where the model is deployed.
You can find the correct answers here
a.Explanation. Azure Machine Learning Studio is a web-based portal that allows you to manage resources in a workspace.
b.Explanation
bExplanation
a.Explanation: To get the default datastore, use the Workspace.get_default_datasetore method.
b. Explanation: To set the compute target, use the compute_target parameter.
c.Explanation : To pass data between steps, use a PipelineData.
c.Explanation. To run the scoring script in parallel, you should use a ParallelRunStep step.
c.Explanation:

Oswald