Welcome to a transformative journey through the world of data, artificial intelligence, and cutting-edge cloud technologies. In this comprehensive course, we'll equip you with the knowledge and skills needed to excel in the ever-evolving realm of data engineering and machine learning on the Azure platform.
Module 1: Explore Data Concepts and Workloads
- What is Data?
- Describe Concepts of Relational Data
- Explore Concepts of Non-Relational Data
- Examine components of a modern data warehouse
- Artificial Intelligence in Azure
- Responsible Artificial Intelligence
- Azure Machine Learning
- Computer Vision Concepts
- Introduction to Continuous Integration & Continuous Deployment with Azure DevOps
- Introduction to Kubernetes
Module 2: Data Engineering and Machine Learning
- What is Azure Data Factory
- Transforming data with the ADF Mapping Data Flow
- Triggering and Monitoring
- What is a Model?
- What are Pipelines?
- Azure Machine Learning Designer
- Register and Manage Datastores with Azure ML
- Create and Manage Datasets
- Managing Azure Databricks
- Azure Databricks Workspaces
- Storage Integration with Azure Databricks
- KeyVault Integration
MLOps Guided Hackathon
Azure Machine Learning enables you to create a workspace with compute instance and use the Azure Machine Learning designer to define a workflow and create a machine learning model that can be published as a web service for inferencing. It's important to be able to interpret machine learning models to understand how they make predictions and explain the rationale for machine learning-based decisions. Unintentional bias can be incorporated into machine learning models, leading to issues with fairness, and it's necessary to consider the potential impact on different groups. Monitoring for data drift and retraining machine learning models as necessary is important to ensure continued accuracy of predictions over time.
- Challenge 1: Create an Azure ML Workspace and Workflow using Diabetes Data
- Challenge 2: Use Azure ML Learning to interpret a model.
- Challenge 3: Detect and Mitigate Unfairness in Models
- Challenge 4: Monitoring for Data drift