Learn Azure Machine Learning for Data Science and AI

Azure Machine Learning is a powerful tool that can help you unlock the full potential of your data. It allows you to build, deploy, and manage machine learning models at scale.

To get started with Azure Machine Learning, you'll need to create a workspace, which is the central hub for all your machine learning activities. This workspace will serve as the foundation for your projects and experiments.

With Azure Machine Learning, you can automate the machine learning process, from data preparation to model deployment. This means you can focus on the creative aspects of machine learning, rather than getting bogged down in tedious tasks.

By using Azure Machine Learning, you can also improve the accuracy and reliability of your models, thanks to its built-in support for model interpretability and explainability.

Azure Machine Learning offers a range of key features and capabilities that make it an attractive choice for machine learning enthusiasts.

Azure Machine Learning has an on-demand compute that can be customized based on the workload, making it easy to scale up or down as needed. It also has a data ingestion engine that accepts a wide variety of sources, making it easy to get started with your project.

Some of the key features include workflow orchestration, machine learning model management, and metrics & logs of all model training activities. These features make it easy to manage and monitor your machine learning projects.

Here are some of the key features and capabilities of Azure Machine Learning:

Azure offers a range of services that make it easier to develop and deploy machine learning models. You can leverage services like Azure App Services and Azure IoT Edge from inside the Azure ML framework.

Azure Machine Learning has some fantastic features, including on-demand compute that you can customize based on your workload. This means you only pay for what you use, which can be a huge cost-saver.

One of the key benefits of Azure AI services is that they offer pre-built APIs for building intelligent applications. These APIs are easy to set up and require minimal machine learning or data science expertise.

Azure AI services provide a wide range of capabilities, including vision, speech, language, and decision APIs. These APIs can be used to develop apps across devices and platforms.

Here are some of the key features of Azure AI services:

Azure Machine Learning also offers workflow orchestration, which makes it incredibly simple to manage your machine learning workflow. This means you can focus on solving problems, not setting up infrastructure.

ML.NET is an open-source, cross-platform machine learning framework that lets you build custom machine learning solutions and integrate them into your .NET applications. It offers varying levels of interoperability with popular frameworks like TensorFlow and ONNX for training and scoring machine learning and deep learning models.

You can use ML.NET for resource-intensive tasks like training image classification models, and take advantage of Azure to train your models in the cloud.

One of the best things about ML.NET is that you don't need a data science and machine learning experience to use it - it's designed to be accessible to developers who are already familiar with .NET.

ML.NET supports two languages: C# and F#. It also covers all three machine learning phases: data preparation, training, and deployment.

Here are some key benefits of using ML.NET:

The Designer Code Features offer a powerful no-code environment, allowing you to accomplish complex tasks without writing a single line of code.

With the designer mode, you can access a wide range of pre-defined modules for data ingestion, feature selection and engineering, model training, and validation.

You can also add custom scripts to a data flow, giving you the flexibility to incorporate your own Python, R, and SQL logic.

The ML designer's explanation generator is a game-changer, allowing you to understand the context and better interpret the results of your model's performance.

This feature is turned off by default to save compute resources, but you can manually turn it on by going into the Parameters setting.

You'll learn how to describe the capabilities of no-code machine learning with Azure Machine Learning Studio. This is a great skill to have, especially if you're new to machine learning.

By taking this course, you'll be able to identify core tasks in creating a machine learning solution, which will help you build a strong foundation in the field. This includes understanding how to describe core machine learning concepts and identifying common machine learning types.

With this knowledge, you'll be able to create and publish models without writing code using Azure Machine Learning. This is a game-changer for anyone looking to get into machine learning without a strong programming background.

Here are some of the key skills you'll learn:

By mastering these skills, you'll be well on your way to a career in machine learning and artificial intelligence. You'll also be preparing yourself for the AI-900: Microsoft Azure AI Fundamentals exam, which is a great way to demonstrate your knowledge and skills to potential employers.

In this course, you'll learn about various types of machine learning that can be applied in different scenarios. You'll discover that there are common machine learning types.

Some of the most popular machine learning types include supervised and unsupervised learning. Supervised learning involves training a model on labeled data, where the correct output is already known. This approach is useful for tasks like image classification and predictive modeling.

Unsupervised learning, on the other hand, involves training a model on unlabeled data, where the correct output is not known. This approach is useful for tasks like clustering and dimensionality reduction.

Here's a summary of the common machine learning types you'll learn about:

Regression is a supervised machine learning technique used to predict numeric values. It's a powerful tool for making predictions based on data.

In Azure Machine Learning, regression models are created using the designer. This involves creating a workspace, which takes just 1 minute to set up.

A regression model requires compute resources to run, which can take up to 10 minutes to create. This is an important step in the process.

To create a regression model, you need to explore your data first. This can take around 15 minutes, depending on the size of your dataset.

Once you've explored your data, you can create and run a training pipeline, which takes 18 minutes to complete. This is where the magic happens, and your model starts to learn.

After training your model, you need to evaluate its performance. This can take around 10 minutes, and it's essential to ensure your model is accurate.

With a trained model, you can create an inference pipeline, which takes 12 minutes to set up. This pipeline allows you to make predictions using your model.

Finally, you can deploy your predictive service, which takes just 8 minutes to complete. This is the final step in creating a regression model using Azure Machine Learning.

A Classification is a type of machine learning that predicts categories or classes, and it's used in supervised learning techniques. This type of machine learning is useful for making predictions based on labeled data.

To create a classification model, you'll need to use Azure Machine Learning designer. This is a great tool for creating classification models because it's easy to use and requires minimal coding knowledge.

In Azure Machine Learning designer, you can create a classification model by following a series of exercises. These exercises include creating an Azure Machine Learning Workspace, creating compute resources, exploring data, creating and running a training pipeline, evaluating a classification model, creating an inference pipeline, deploying a predictive service, and cleaning up.

Here's a breakdown of the exercises involved in creating a classification model:

Clustering is an unsupervised machine learning technique used to group similar entities based on their features. It's a powerful tool for identifying patterns and relationships in data.

In Azure Machine Learning, clustering models can be created using the designer, which is a user-friendly interface for building and deploying machine learning models. The process involves creating a Microsoft Azure Machine Learning workspace, which can be done in just 8 minutes.

To create a clustering model, you'll need to create compute resources, which can take around 10 minutes. This step is crucial for running the training pipeline, which takes 12 minutes to complete.

Here's a brief overview of the clustering process:

By following these exercises, you'll be able to create a clustering model and deploy a predictive service in Azure Machine Learning. The entire process, from creating a workspace to deploying a service, can be completed in under an hour.

An Azure Workspace is a centralized place to manage resources required for training a model, including computing targets, data for model training, Notebooks, Experiments, Models, Pipelines, etc.

To access these resources, users need to authenticate using the Azure Active directory. This is a crucial step in using Azure Workspace effectively.

A key feature of Azure Workspace is the ability to create multiple compute instances, which are online computed resources that already have a development environment installed to write and run code in Python.

Here are some of the development platforms and tools available for machine learning:

The Azure Data Science Virtual Machine is a customized virtual machine environment on the Microsoft Azure cloud. It's available in versions for both Windows and Linux Ubuntu.

This environment is built specifically for doing data science and developing machine learning solutions. It has many popular data science, machine learning frameworks, and other tools pre-installed and pre-configured to jump-start building intelligent applications for advanced analytics.

The latest versions of all commonly used tools and frameworks are included in the Data Science VM. You can also use virtual machine options with highly scalable images that have graphics processing unit (GPU) capabilities for intensive data modeling.

Here are some key benefits of using the Azure Data Science Virtual Machine:

However, keep in mind that the virtual machine cannot be accessed when offline, and running a virtual machine incurs Azure charges.

SQL is a powerful tool in the Azure Machine Learning Workspace and Tools ecosystem. It allows you to add statistical analysis, data visualization, and predictive analytics to your relational data in Python and R.

You can use SQL machine learning on-premises or in the cloud, making it a flexible choice. Current platforms and tools include SQL Server Machine Learning Services, Azure SQL Managed Instance Machine Learning Services, Machine learning in Azure Synapse Analytics, and Machine Learning extension for Azure Data Studio.

SQL machine learning is perfect for encapsulating predictive logic in a database function, making it easy to include in data-tier logic.

Here are some key details about SQL machine learning:

Automated Machine Learning is a game-changer for building predictive models. It can save you time and compute resources by automating the process of training and deploying machine learning models.

With Automated Machine Learning in Azure Machine Learning, you can identify different kinds of machine learning models and use the automated machine learning capability to train and deploy a predictive model. This can be a huge time-saver, especially in the iterative process of building machine learning models.

The Azure Machine Learning Studio supports only supervised machine learning models where you have training data and known labels. These models include classification, regression, and time series forecasting.

Automated Machine Learning can help you optimize algorithms for the best outcomes with the compute power that comes with Azure. This takes away the need for manual trial and error iterations that come with building a model.

To run an automated machine learning algorithm, you need to specify the dataset with labels, configure the automated machine learning run, select the algorithm and settings, and review the best model generated. This process can help you identify the best model for a particular use case.

Here are some key metrics you can configure in the studio:

You can use Azure Machine Learning to manage big data by utilizing services like Azure SQL Database, Azure Cosmos DB, and Azure Data Lake to ingest and process large volumes of data.

To create a training script, you'll need to create a folder to save your Python scripts, and then use the Run feature to monitor the trial, log metrics, and store the output of the trial.

When creating a training script, it's essential to create datasets X and Y, where X contains feature variables and Y contains the output variable. You'll also need to split the dataset into train and test sets, typically in a 70:30 ratio.

The training script should include steps to save the trained model to a desired folder. In Azure Machine Learning, you can use the get_metric method of the run class to print metrics like Regularization rate, AUC, and Accuracy.

Data processing capabilities in Azure Machine Learning Studio include imputing missing values, encoding categorical features, and balancing data by normalizing and scaling features. You can also derive time-series features by extracting day, month, and year from the date field.

Here are some key steps to follow when creating a training script:

By following these steps, you'll be able to create a training script that effectively trains your machine learning model and prepares it for deployment.

Creating an inference pipeline is a crucial step in Azure Machine Learning. You'll have two options to choose from.

One option is a real-time inference pipeline, which allows you to make predictions in real-time with an immediate response from the service. This is ideal for applications where speed and timeliness are critical.

The other option is a batch inference pipeline, which stores predictions as files for business applications. This is suitable for scenarios where data needs to be processed in bulk.

Consider the specific needs of your project when deciding between these two options.

To get started with Azure Machine Learning, you'll need to create an Azure account, which is free for the first 12 months with 13,300 credits. This will give you a chance to practice and get familiar with the platform.

Azure ML Studio is a great place to start, as it's convenient to use and offers a drag-and-drop interface for building Machine Learning models. You can select a dataset from the Saved Datasets menu and use the two-class regression algorithm for training.

To begin, select the dataset you want to work with and launch the column selector to choose the relevant columns. Then, search for Split Data and drop it on the workspace to split your data into training and testing sets.

Once you have your data split, search for Train Model and drop it on the workspace. Connect the Train Model to the Split Data and drag and drop Evaluate Model to evaluate the performance of your model.

Some examples of anomaly detection in Azure ML Studio include identifying and predicting rare data points, such as fraud detection or abnormal equipment readings.