The course, Machine Learning for Beginners , is designed to empower students of all ages to learn the basics of ML, says the company. It presumes no knowledge of ML, and is offered as a free 12-week, 24-lesson curriculum, plus a bonus 'postscript' lesson.
"Travel around the world in this themed semester-long self-study course," says Microsoft Cloud Developer Advocate Chris Noring, "as we look at ML topics through the lens of world cultures."
The curriculum is all about "classic Machine Learning," so its basic concepts are primarily covered using Scikit-learn, a Python ML library that helps demystify and explain these concepts. Deep learning and neural networks are not covered in this ML curriculum, says the company, but will be in an upcoming "AI for Beginners" curriculum this fall.
The ML course's curricula are structured with a modified Project-Based pedagogy and include:
- a pre-lesson warmup quiz
- a written lesson
- knowledge checks
- a project to build
- infographics, sketchnotes, and visuals
- a challenge
- an assignment
- a post-lesson quiz
- a 'PAT' (see below)
- opportunities to deepen knowledge on Microsoft Learn
The lessons are grouped so that users can deep-dive into various important aspects of classic ML. The course starts with an introduction to ML concepts, moving to its history, concepts of fairness in machine learning, and a discussion of the tools and techniques of the trade.
It then moves on to Regression, Classification, Clustering, Natural Language Processing, Time Series Forecasting, Reinforcement Learning, with two 'applied' lessons demonstrating how to use ML models within web apps for inference. A 'postscript' lesson lists "real-world" applications of ML, showing how these techniques are used "in the wild."
To make it easy for new learners to get started with ML, the company says it built the content so that it can be used offline and so that the exercises can be completed using .ipynb notebooks within Visual Studio Code.
Microsoft acquisition aims to bring