AI-900 Day 8 Fundamentals of Machine Learning in Azure

What is the primary purpose of Azure Machine Learning?

• To create and manage virtual machines

• To build, train, and deploy machine learning models

• To store large datasets

• To develop web applications

Answer explanation

The primary purpose of Azure Machine Learning is to build, train, and deploy machine learning models, enabling users to create intelligent applications and automate processes effectively.

Which of the following is a key component of Azure Machine Learning?

• Azure Functions

• Azure Blob Storage

• Azure Machine Learning Studio

• Azure SQL Database

Answer explanation

Azure Machine Learning Studio is a key component that provides a web-based interface for building, training, and deploying machine learning models, making it essential for users working with Azure Machine Learning.

What is the first step in the data preparation process?

• Data cleaning

• Data transformation

• Data collection

• Data splitting

Answer explanation

The first step in the data preparation process is data collection. This involves gathering the necessary data from various sources before any cleaning, transformation, or splitting can occur.

Which technique is used to handle missing data in a dataset?

• Data augmentation

• Data imputation

• Data normalization

• Data encryption

Answer explanation

Data imputation is the technique used to fill in missing values in a dataset, making it essential for data analysis. Other options like data augmentation, normalization, and encryption serve different purposes.

What is the purpose of data normalization?

• To encrypt data for security

• To scale data to a standard range

• To remove duplicates from data

• To split data into training and test sets

Answer explanation

The purpose of data normalization is to scale data to a standard range, which helps improve the performance of machine learning algorithms by ensuring that no single feature dominates due to its scale.

Which of the following is a common metric for evaluating classification models?

• Mean Absolute Error (MAE)

• Root Mean Square Error (RMSE)

• Accuracy

• R-squared

Answer explanation

Accuracy is a common metric for evaluating classification models, as it measures the proportion of correctly classified instances. In contrast, MAE, RMSE, and R-squared are metrics used for regression models.

What is overfitting in machine learning?

• When a model performs well on training data but poorly on new data

• When a model performs well on both training and new data

• When a model performs poorly on both training and new data

• When a model performs well on new data but poorly on training data

Answer explanation

Overfitting occurs when a model learns the training data too well, capturing noise and details that do not generalize to new data. This results in high accuracy on training data but poor performance on unseen data.

Which of the following is a method to prevent overfitting?

• Increasing the learning rate

• Reducing the dataset size

• Using regularization techniques

• Decreasing the number of features

Answer explanation

Using regularization techniques helps to prevent overfitting by adding a penalty for larger coefficients in the model, thus promoting simpler models that generalize better to unseen data.

What is the purpose of a validation set in model training?

• To train the model

• To test the model

• To tune model parameters

• To deploy the model

Answer explanation

The validation set is used to tune model parameters by evaluating the model's performance on unseen data during training. This helps prevent overfitting and ensures the model generalizes well to new data.

Which of the following is a common method for deploying machine learning models in Azure?

• Azure Logic Apps

• Azure Kubernetes Service

• Azure DevOps

• Azure Virtual Machines

Answer explanation

Azure Kubernetes Service (AKS) is a common method for deploying machine learning models in Azure, as it provides a scalable and managed environment for containerized applications, making it ideal for ML workloads.