Google, IBM & Meta Certificates – 40% Off
One Coursera Plus subscription covers most Professional Certificates on Coursera.
Unlock All Certificates
- Choose this module if you want to design a machine learning training solution and select Azure Machine Learning services and compute.
In this module, you learn how to:
- Design a solution to get and prepare data.
- Choose a service and compute to train a model.
- Prepare for model deployment options.
- Choose this module if you want to compare models by using automated machine learning, MLflow, and the Responsible AI dashboard.
In this module, you learn how to:
- Prepare your data to use AutoML for classification.
- Configure and run an AutoML experiment.
- Evaluate and compare AutoML models.
- Configure MLflow for model tracking in notebooks.
- Use MLflow for model tracking in notebooks.
- Evaluate a trained model using the Responsible AI dashboard.
- Choose this module if you want to convert notebooks to training scripts, run command jobs, and track models with MLflow.
In this module, you'll learn how to:
- Convert a notebook to a script.
- Test scripts in a terminal.
- Run a script as a command job.
- Use parameters in a command job.
- Use MLflow when you run a script as a job.
- Review metrics, parameters, artifacts, and models from a run.
- Choose this module if you want to optimize model training by tuning hyperparameters with sweep jobs in Azure Machine Learning.
In this module, you learn how to:
- Define a hyperparameter search space.
- Configure hyperparameter sampling.
- Select an early-termination policy.
- Run a sweep job.
- Choose this module if you want to automate multistep machine learning workflows with reusable components and Azure Machine Learning pipelines.
In this module, you learn how to:
- Create components.
- Build an Azure Machine Learning pipeline.
- Run an Azure Machine Learning pipeline.
- Choose this module if you want to automate and validate model training with GitHub Actions and Azure Machine Learning pipelines.
In this module, you learn how to:
- Identify which machine learning assets to version in Git and which belong in purpose-built services such as data stores and model registries.
- Describe trunk-based development practices — short-lived branches, pull requests, and branch protection rules — that keep the shared branch stable while allowing safe iteration.
- Explain how GitHub Actions workflows validate code changes with linting and unit tests, and how branch protection can require those checks before merging.
- Compare service principal client secrets with workload identity federation, and explain why OpenID Connect is the preferred authentication approach for GitHub Actions workflows that call Azure.
- Describe how a GitHub Actions workflow submits Azure Machine Learning command and pipeline jobs, and explain how an intermediary converts Azure events into repository_dispatch events.
- Choose this module if you want to register, deploy, monitor, and roll back models by using GitHub Actions and protected environments.
After completing this module, you can:
- Register, version, and archive MLflow models throughout their lifecycle.
- Use GitHub environments to control access and require approval before production promotion.
- Deploy and troubleshoot a managed online endpoint, safely promote a model, and roll back to a previous version.
- Automate model deployment and testing with GitHub Actions.
- Monitor a deployed model and recognize signals that call for investigation or retraining.
