Most data science projects fail. There are various reasons why, but one of the primary reasons is the challenge of deployment. One piece to the deployment puzzle is understanding how to automate your pipeline’s functions and continuously optimize its performance, which is why we developed this course, MLOps2: Data Pipeline Automation & Optimization using Microsoft Azure Machine Learning (AML).
Class Deals by MOOC List - Click here and see EdX's Active Discounts, Deals, and Promo Codes.
Most data science projects fail. There are various reasons why, but one of the primary reasons is the challenge of deployment. One piece to the deployment puzzle is understanding how to automate your pipeline’s functions and continuously optimize its performance, which is why we developed this course, MLOps2: Data Pipeline Automation & Optimization using Microsoft Azure Machine Learning (AML). In this course you will learn how to set up automated monitoring of your data pipeline for prediction. Data drift, model drift and feedback loops can impair model performance and model stability, and you will learn how to monitor for those phenomena. You will also learn about setting triggers and alarms, so that operators can deal with problems with model instability. You will also cover ethical issues in machine learning and the risks they pose, and learn about the "Responsible Data Science" framework.
This course is part of the Machine Learning Operations with Microsoft Azure (MLOps with Azure) Professional Certificate.
What you'll learn
You will learn how to set up automated monitoring of your data pipeline for prediction and get hands on experience with topics like data pipelines, drift and feedback loops, model stability, triggers & alarms, model security, responsible AI and much more.
But most importantly, by the end of this course, you will know…
- How to meet the differing requirements of model training versus model inference in your pipeline
- How to check for model drift, data drift, and feedback loops
- How to apply the principles of Continuous Integration (CI), Continuous Delivery (CDE) and Continuous Deployment (CD)
Syllabus
Week 1 – Drift and Feedback Loops
Module 1: Training Versus Inference Pipelines
Module 2: Drift & Feedback Loops
Week 2 – Triggers, Alarms & Model Stability
Module 3: Triggers & Alarms
Module 4: Model Stability
Week 3 – CI/CD (Continuous Integration & Continuous Deployment/Delivery)
Module 5: CI/CD
Week 4 – Model Security and Responsible AI
Module 6: Responsible AI
Prerequisites:
Participants should have taken the first two courses:
- Predictive Analytics: Basic Modeling Techniques
- Machine Learning Operations (MLOps 1): Deploying AI and ML Models in Production using AML
and be comfortable working with Python in a cloud-based environment. Learners will gain maximum benefit if they have some familiarity with software development, including git, logging, testing, debugging, code optimization and security.
