Azure Machine Learning (AML) Training in Astana, Kazakhstan

This course introduces participants to Azure Machine Learning (Azure ML), Microsoft's cloud-based machine learning service. It covers the end-to-end ML lifecycle including data preparation, model training, evaluation, and deployment using Azure ML Studio and SDKs. The course balances conceptual learning with hands-on labs to provide practical exposure to building scalable ML workflows in the cloud.

Course Agenda

Introduction to Azure Machine Learning

• What is Azure ML?

• Use cases and architecture

• Azure ML Studio vs. SDK

Setting Up Your Environment

• Creating an Azure ML workspace

• Exploring the Azure ML Studio interface

• Using Azure Notebooks and VS Code integration

Data Preparation and Management

• Connecting to data stores

• Data import, labeling, and versioning

• Data exploration and cleansing

Model Building and Training

• Creating experiments

• Training models with AutoML

• Custom training with Python and Scikit-learn

• Hyperparameter tuning

Model Evaluation

• Evaluating metrics: accuracy, precision, recall, F1, AUC

• Visualizing results

• Model explainability and fairness analysis

Deployment and Operationalization

• Deploying models to AKS or ACI

• Consuming models via REST APIs

• Model versioning and rollback

Monitoring and Retraining

• Monitoring endpoints and metrics

• Setting up alerts

• Retraining workflows and automation

Governance and Cost Management

• Managing compute resources

• Security, roles, and access control

• Estimating and optimizing costs in Azure ML

Hands-on Exercises