U. Michigan Online Course Computational ML for Scientists and Engineers

The University of Michigan’s Continuum Jumpstart Course Computational Machine Learning (ML) for Scientists and Engineers is designed to equip you with the knowledge you need to understand, train, design and machine learning algorithms, particularly deep neural networks, and even deploy them on the cloud.

You’ll learn by programming machine learning algorithms from scratch in a hands-on manner using a one-of-a-kind cloud-based interactive computational textbook that will guide you, and check your progress, step-by-step. Using real-world datasets and datasets of your choosing, you will understand, and we will discuss, via computational discovery and critical reasoning, the strengths and limitations of the algorithms and how they can or cannot be overcome. You will understand how machine learning algorithms do what they claim to do so you can reproduce these while being able to reason about and spot wild, unsupported claims of their efficacy.

By the end of the course, you will be ready to harness the power of machine learning in your daily job and prototype, we hope, innovative new ML applications for your company with datasets you alone have access to.

It’s ideal for folks who want to go deeper than a regular MOOC and want to learn by coding.

We’ll use Julia from start to finish and sprinkle in Python (via Keras and PyTorch) so you can become bilingual and also appreciate how much nicer Julia is! (if you don’t already know)

See Jumpstart-ML for a description – apply by Jan 8th, 2021 for cohort starting Feb 15th.

See Testimonials and Advice

for testimonials from the pilot cohort.


:+1: As someone who’s currently taking this course (obvious disclaimer, and I’m still not through!) - I cannot recommend this course highly enough. I am a finance professional who wanted to upskill into Machine Learning. This course is definitely a cut above the usual crop of machine learning courses I found online - no hand-wavy explanations, building the first models by hand (SGD / Nesterov).

And its in Julia!

Only warning: It is a lot of hard work, so be prepared for that.


Thanks, Gurvesh for the recommendation. It’s been a pleasure interacting with you via the course and inspiring to see how deeply you are diving in!

Btw, Gurvesh is taking the harder version that also involves Computational Linear Algebra (from scratch) via Applied Computational Linear Algebra for Everyone | Mynerva

The Comp ML course does need work because linking math to code gets tricky and everyone gets stuck somewhere. You’ll be doing and learning (way) more than a “run all” style Jupyter/CoCal notebook. You’ll build up the code bit by bit using the mynerva.io codex approach described in this JuliaCon talk that has baked in step-by-step checking. When you get stuck, or have more questions, we are happy and want to help you get unstuck because that’s where the real (human) learning happens and where there is an opportunity for a deeper conversation.


We just started enrollment in the Fall cohort starting August 11, 2021.



for application and


for some testimonials.

The course is heavily Julia based :slight_smile:

1 Like

I was sold as soon as I saw @travigd was involved!


Yup :slight_smile: can’t do it without mynerva.io (soon to be pathbird)

1 Like