Linear Algebra Course

[trshstar]

Description

I would like to create a course on Linear Algebra, maybe like linear algebra basics notebook that goes over the important concepts in linear algebra needed for machine learning basics, and how they are applied in machine learning, want to make use of visualization libraries as some concepts in linear algebra can be abstract.

Let me know if there are specific concepts you would like me to go over!

[Haleshot]

Hi Kevin! Thanks a lot for proposing a new topic (and apologies for the slightly delayed reply).

In relation to the course, I had actually established a collaboration with Deep-ML and GroundZeroAI for the creation of a Linear Algebra course a while back.

• You can find the short series of Linear Algebra notebooks made here: https://marimo.io/c/@haleshot/linear-algebra
• The relevant YouTube Video lectures (from GroundZeroAI's YT channel) for them here: Playlist

linear algebra needed for machine learning basics, and how they are applied in machine learning, want to make use of visualization libraries as some concepts in linear algebra can be abstract.

I'm curious to see some cool notebooks topics being proposed for this. Would you be able to draw up a tentative proposed outline (checkout other issues for the template/format) for this course? Feel free to list references too.

Let me know if there are specific concepts you would like me to go over!

Off the top of my head, can think of SVD/Eigenvalues/vector based usecases (face recognition, etc.). Checkout this cool notebook on SVD that was one-shotted via marimo AI!

[tusharrrr1]

If this issue still persist maybe i can work on it

[Haleshot]

If this issue still persist maybe i can work on it

Hi Tushar! Thanks for expressing interest! Do you have any tentative proposed topics in mind (as stated in my comment above)? Let me know what you think after referring to the resources stated for Linear Algebra and marimo above.

Moreoever, you might find two videos on Linear algebra that came out recently using marimo relevant:

• The Matrix as a User Interface: https://youtu.be/dVrkzc_arE8?si=Ih_TXN_R1rG_aC_B
• Previewing some Linear Algebra: https://youtu.be/1SJOqnZTV-E?si=KqWzNpFa5bROMNHV

[tusharrrr1]

Hi @Haleshot, thanks for sharing the resources 🙌. I went through the notebooks and videos you linked — really insightful! Based on that, here’s a tentative outline I’d propose for a Linear Algebra Basics for Machine Learning course (using marimo + visualizations):
Tentative topics:

1. Visualization-first notebooks – making heavy use of interactive plots (e.g., showing transformations in real-time, sliders to change eigenvalues, etc.)
2. ML application mini-projects – rather than just covering the math, build small end-to-end demos

This could be organized as a progressive set of notebooks where each topic builds toward applying linear algebra concepts in ML.

[Haleshot]

@tusharrrr1 That does sound nice! I like the Eigen values part and think something similar can be built for it (like it was done for SVD one whose links I posted above in this thread). Considering PCA (dimensionality reduction) and image recognition are some great usecases/practical applications of eigen values/faces, would love to see practical demos being made as well (in addition to theory-specific ones).

Same with the mini-projects part; I do believe you could use our AI notebook generation website to one shot (and later refine) some cool demos (small ones) using LA.

Any thoughts on the above? @trshstar

[tusharrrr1]

Hi @Haleshot, thanks for the feedback! 🙌

I really like your suggestion of building practical demos around eigenvalues/PCA and face recognition — that definitely makes the concepts more concrete

If you assign me this for parallel contribution I can start by drafting the first couple of notebooks (foundations + transformations) and then move into eigenvalues/PCA with demos. Would love your input on whether this structure feels aligned with what you had in mind!

[Haleshot]

The first set of notebooks planned do sound great! The ideas I had are mostly random topics that I suggested above; nothing as a proposed structure for a course/series of notebooks as such. Would be great to go over your notebook contribs/PRs once you push them here!

[trshstar]

Hey sorry for the late response, beginning of the semester has been hectic. Sounds like there is kind of an outline already, I like the ideas, very practical real world use cases. I can pick up whatever @tusharrrr1 leaves off at, or if you want to split the work let me know.