INTRODUCTION – Exam Preparation Course 2
In this module, you will be reading a thorough review of Course 2, part of the Microsoft Azure Data Scientist Associate specialization. This course is an extension of the knowledge derived from Course 1, as it gives you capabilities to become a data scientist or machine-learning personifier using well-designed and effective Azure tools and services. You would revisit aspects such as advanced data preparation techniques, feature engineering, and model selection. The course also addresses automating the entire machine learning lifecycle-from data ingestion and preprocessing to model training and evaluation and, ultimately, deployment-of machine learning projects using Azure Machine Learning.
Further, the course covers best practices to be followed for monitoring and managing an ML model in production to ensure consistent performance and reliability. This review will strengthen your understanding of these advanced topics while equipping you with practical skills to solve real-life challenges in data science using Azure.
Learning Objectives
- Define the major course contents within the Microsoft Azure Data Scientist Associate specialization
Summarize important topics covered in Course 2 - Create no-code predictive models with Azure Machine Learning
- Evaluate one’s knowledge and competencies in creating no-code predictive models using Azure Machine Learning.
Quiz: Create no-code predictive models with Azure Machine Learning
1. What data values are influencing prediction models?
- Labels
- Identifiers
- Features (CORRECT)
- Dependent variables
Correct: Prediction models are typically influenced by the data attributions or variables called features. This means that input data is put into machine learning algorithms for their predictions or classifications.
2. Let’s suppose you want to create an AI system that can predict how many minutes late a flight will arrive based on the amount of snowfall at an airport. Which machine learning type should you use?
- Classification
- Regression (CORRECT)
- Clustering
Correct: Regression is a method in supervised machine learning that can be used to predict soft continuous numbers according to input features. In essence, it creates a relation between the independent variables (features) and the dependent variables (target) for the prediction.
3. Imagine you work for a government institution that wants to predict the sea level in meters for the following 10 years. Which type of machine learning should you use?
- Regression (CORRECT)
- Clustering
- Classification
Correct: Regression is a type of supervised machine learning technique that helps to estimate values by predicting the continuous nature of these values.
4. Azure Machine Learning designer lets you visually connect datasets and modules on an interactive canvas to create machine learning models. Which two components can be dragged-and-dropped onto the canvas?
Select all options that apply.
- Pipeline
- Compute
- Dataset (CORRECT)
- Module (CORRECT)
Correct: Datasets and modules are the two elements which can be dragged onto the canvas.
Correct: Datasets and modules-two entities-active in the world of machine learning or data science tools like Azure Machine Learning or other-are the two most easily dragged and dropped onto the canvas.
5. True or False?
When working in Azure Machine Learning designer, it is possible to save your progress as a pipeline draft.
- True (CORRECT)
- False
Correct: With Azure Machine Learning Designer, you could save your work as a draft pipeline.
6. You can use AI systems to predict whether a student will complete a university course. Which machine learning type enables you to do that?
- Classification (CORRECT)
- Regression
- Clustering
Correct: Classification is a supervised machine learning technique used to predict or identify categories or labels.
7. True or False?
Accuracy is always the primary metric used to measure a model’s performance. Is this true?
- True
- False (CORRECT)
Correct: There are different metrics that can be used to measure a model’s performance.
8. True or False?
Automated machine learning can automatically infer the training data from the use case provided.
- True
- False (CORRECT)
Correct: Automated machine learning cannot autonomously infer the training data from the present use case.
9. True or False?
Azure Machine Learning designer provides a drag-and-drop visual canvas to build, test, and deploy machine learning models.
- True (CORRECT)
- False
Correct: For the whole visualization of the process of building, testing, and deploying ML models, Azure Machine Learning has a drag-and-drop canvas feature.
10. Fill in the blank.
__________ is a form of machine learning that has the capability to group similar items based on their features.
- Regression
- Clustering (CORRECT)
- Classification
Correct: Clustering is an unsupervised machine-learning approach in which similar entities are grouped together according to their properties.
11. In a machine learning algorithm, what method should you use to split data for training and evaluation?
- Randomly split the data into rows for training and columns for evaluation
- Use labels for training and features for evaluation
- Use features for training and labels for evaluation
- Randomly split the data into rows for training and rows for evaluation (CORRECT)
Correct: The split by percentage is a way to split the data in Azure Machine Learning. Through this method, a percentage of the data is randomly divided into training and testing sets.
12. Which of the following metrics is used to evaluate a classification model?
- Coefficient of determination (R2)
- Root mean squared error (RMSE)
- True positive rate (CORRECT)
- Mean absolute error (MAE)
Correct: True positive rate is the ideal metric in classifying models.
13. Which module in the Azure Machine Learning designer should you use if you want to create a training dataset and a validation dataset from an existing dataset?
- Select columns in dataset
- Split data (CORRECT)
- Join data
- Add rows
Correct: The data can be divided into training and validation sets by splitting them.
14. Let’s suppose you are working on an AI application that should predict the weather. From the dataset you have, you want to pick temperature and pressure to train the model. Which machine learning task enables you to do that?
- Feature engineering
- Feature selection (CORRECT)
- Model training
Correct: The selection of features is a process whereby a disciplined subset of the relevant features for the design of an analytical model is chosen to be included in the design.
15. Predicting whether someone uses a bicycle to travel to work based on the distance from home to work is a use case for?
- Clustering
- Regression
- Classification (CORRECT)
Correct: Classification is characterized as a supervised form of machine learning applicable for predicting discrete categories or classes.
16. You want to create a CRM application that uses AI to segment customers into different groups to support a marketing department. Which machine learning type should you use?
- Classification
- Clustering (CORRECT)
- Regression
Correct: Clustering can be regarded as an unsupervised machine learning technique. It is employed to cluster or segregate similar entities grouped in clusters based on their characteristics or features.
17. True or False?
Azure Machine Learning designer supports custom JavaScript functions.
- True
- False (CORRECT)
Correct: Custom JavaScript functions are not supported in Azure Machine Learning Designer.
18. Predicting how many minutes it will take someone to run a race based on past race times is a use case for?
- Regression (CORRECT)
- Clustering
- Classification
Correct: Regression is one of the supervised types of machine learning which is basically used to predict continuous numeric values.
CONCLUSION – Exam Preparation Course 2
This module will enable you to have an in-depth understanding of advanced data science and machine learning techniques using Azure, as it appears in Course 2 of the Microsoft Azure Data Scientist Associate specialization. You will acquire practical skills in advanced data preparation, feature engineering, selecting models, and automating the machine learning lifecycle with Azure Machine Learning.
You will also be equipped with best practices on monitoring and managing models in production for these to continue to perform reliably. Thus, you will have a strong background for tackling real-world data science challenges and advancing in your specialization track with comprehensive knowledge.