📄️ Azure Machine Learning
Build, train and ship a demand-forecasting model on Azure ML v2: workspace, compute, data assets, environments, jobs, AutoML, registry, endpoints, pipelines and cost control.
📄️ 1. Workspace and resources
Module 1 of the Azure Machine Learning premium course: what the workspace really is, its four attached resources, and the identities and roles that make it usable.
📄️ 2. Compute targets
Module 2 of the Azure Machine Learning premium course: compute instances, auto-scaling clusters, serverless jobs, low-priority nodes, and region quotas.
📄️ 3. Data assets and datastores
Module 3 of the Azure Machine Learning premium course: datastores for Blob and ADLS, versioned data assets (uri_file, uri_folder, MLTable) and identity-based access.
📄️ 4. Environments and images
Module 4 of the Azure Machine Learning premium course: curated versus custom environments, conda files and Dockerfiles, image builds, image cache and version pinning.
📄️ 5. Training jobs and tracking
Module 5 of the Azure Machine Learning premium course: command jobs, inputs and outputs, native MLflow logging, hyperparameter sweeps and reading results in Studio.
📄️ 6. AutoML: uses and limits
Module 6 of the Azure Machine Learning premium course: AutoML forecasting on retail sales, how it works, when it wins over a hand-built baseline, and its real cost.
📄️ 7. Model registry
Module 7 of the Azure Machine Learning premium course: registering MLflow and custom models, versions, tags, cross-workspace sharing via registries, and full lineage.
📄️ 8. Online and batch endpoints
Module 8 of the Azure Machine Learning premium course: managed online endpoints, blue-green deployments and traffic splitting, scoring scripts, and batch endpoints on clusters.
📄️ 9. Pipelines and scheduling
Module 9 of the Azure Machine Learning premium course: reusable components, YAML pipelines, weekly scheduling with cron and recurrence triggers, event-based triggers on new data.
📄️ 10. Monitoring and costs
Module 10 of the Azure Machine Learning premium course: model monitoring for drift, Azure Monitor for endpoints, cores quotas per region, cost analysis by tag and auto-shutdown.
📄️ Recap and exam
Complete recap of the Azure Machine Learning premium course: workspace, compute, data, environments, jobs, AutoML, registry, endpoints, pipelines and monitoring, then the 40-question exam.