Evaluating Model Effectiveness in Microsoft Azure

Question 1

What must you register in order to create an Automated ML run?

Datastore

Data science VM

Notebook VM

Dataset

Answer 1

Dataset

Question 2

Which of Microsoft’s Responsible AI principles concerns humans maintaining control of their AI environment?

Inclusiveness

Accountability

Transparency

Fairness

Answer 2

Accountability

Question 3

Which of the following situations is likely to involve unintended bias?

Non-standard data formats

Large training dataset

Small number of compute nodes

Incomplete training data

Answer 3

Incomplete training data

Question 4

What is the term used to signify the runtime environment offered by a Jupyter notebook?

Notebook VM

Kernel

Script

Model

Answer 4

Kernel

Question 5

Which of the following terms involves prediction generation?

Modeling

Evaluating

Computing

Scoring

Answer 5

Scoring

Question 6

Cross-validation involves splitting a dataset into training and what kind of subset?

Regression

Testing

Compute

Endpoint

Answer 6

Testing

Question 7

What is the name of the draggable elements in the Azure Machine Learning designer?

Package

Library

Class

Module

Answer 7

Module

Question 8

Which Azure Machine Learning designer control enables you to view the underlying source data in a pipeline?

Visualize

Deploy

Explain

Regression

Answer 8

Visualize

Question 9

Overfitting is indicated when the model performs better on which type of data?

Model

Test

Training

Deployment

Answer 9

Training

Question 10

Which ML term describes combining several base models into a single optimal model?

Overfitting

Regression

Classification

Ensemble

Answer 10

Ensemble