TL;DR
Video Series
The "Essence of Linear Algebra" series by 3Blue1Brown is highly recommended for building intuitive understanding through visual explanations. This YouTube playlist is praised for making complex concepts more accessible [1:1],
[2:2],
[4:1]. Additionally, Kimberly Brehm’s lecture series on YouTube has been noted as a useful resource
[4:3].
Textbooks and Courses
Gilbert Strang's lectures on MIT OpenCourseWare (OCW) are frequently mentioned as a solid foundation for learning linear algebra. His textbook complements the video lectures well [1:2],
[4:4]. For those looking for an alternative, "Linear Algebra Done Right" by Sheldon Axler and "Elementary Linear Algebra" by Howard Anton are recommended
[2:4],
[2:5].
Online Platforms
Coursera offers a "Mathematics for Machine Learning" specialization which includes linear algebra content tailored for machine learning applications [5]. Additionally, http://studybyyourself.com provides a free online course with illustrations that can be helpful for beginners
[3:1].
Advanced Resources
For those seeking more advanced material, Halmos's "Finite Dimensional Vector Spaces" is suggested, though it requires familiarity with mathematical proofs [2:3]. Bhatia's "Matrix Analysis" is another advanced option but is not suitable for beginners
[2:3].
Learning Strategies
Scott Young's framework on how to teach yourself math and the ADEPT method from BetterExplained are strategies that can aid in understanding difficult concepts [5:2]. These resources focus on developing insights and intuitive understanding, which can be particularly beneficial when tackling abstract topics like linear algebra.
Basically how can I pick up the basics of Linear Algebra on my own. I don't need to become an expert in the subject, I just want to learn enough that I can apply it to computer science and maybe continue to build on my knowledge in the future.
At this point I know very little about the subject: I know the basics of vector arithmetic, scalar multiplication, etc. so I'm more or less starting from scratch.
Lay's Linear Algebra and its Applications is a pretty good text (I used it for my first two courses) and there's a free pdf download online.
If you want alternate presentations of things in that book, I would check out the 18.06 lecture series on MIT OCW. You can get a full playlist here: youtube.com/watch?v=7UJ4CFRGd-U&list=PL221E2BBF13BECF6C.
This series is pretty great, won’t be everything you need but it’s a great reference and is really helpful to understand what linear algebra actually is. Essence of Linear Algebra
I watched some of that one already and 3B1B is always awesome! I don't mean to sound unappreciative but I think I need something less high level.
Lol no worries! I’m in linear algebra right now and am doing well and take all my notes on my iPad and have classmates that are struggling so I’m going to be making a google drive folder to share with them, I wouldn’t mind sharing with you as well if you’re interested!
Also btw I’m studying comp sci too!
Hey everyone! I hoping to learn linear algebra from scratch to advanced also with its applications in industry using matlab or wolfram. Any resources which would help me with this ? Ps : I’ve started gilbert strang’s lectures on yt
The 3Blue1Brown linear algebra series is excellent for building up a geometrical intuition for things
If you want a much more advanced book than Strang, get a hold of Halmos's "Finite dimensional vector spaces", but make sure you know what a mathematical proof is, and how to come up with them yourselves.
There are much more advanced treatments of linear algebra than Halmos (for example Bhatia's "Matrix Analysis" book). However, that is deep stuff, and not for beginners.
Also read strang’s textbook. Best yet also read linear algebra done right by alxer.
Elementary Linear Algebra by Howard Anton is really good
Besides the OCW lectures, Strang has a book too, I highly recommend getting familiar with reading maths.
3Blue1Brown's videos are phenomenal for building your intuition around linear algebra.
MATLAB/Wolfram: I'd look on YouTube. Or the official courses (example).
Does anyone know of any good online resources for learning linear algebra (web series, practice problems, etc)? I’m taking it right now in college and the professor isn’t very good, so I’m hoping to find some resources that will help me understand what we’re learning in class!
Linear algebra is a vast subject. Are you just doing matrices and solving linear systems? Or vector spaces? Tensor products? We can help better if you provide more detail
I've always enjoyed 3Blue1Brown's YouTube content. He is very good at creating visual arguments for what is being explained.
Here a nice course for beginners. For free. Nicely illustrated. http://studybyyourself.com/seminar/linear-algebra/course/?lang=en.
I've always been interested in machine learning but understanding it beyond the most basic 3 layer neural net will require linear algebra. I took it in university as part of my computer science degree but it was a struggle, so hopefully things have gotten much better, especially with videos and illustrations.
There is the default, Khan Academy, normally a great choice, but wondering if you know of any other good resources?
If you have prior knowladge and you wish to refresh and strenthen it I would sugest this playlist:
​
it also comes with lecturenotes:
https://drive.google.com/file/d/1HSUT7UMSzIWuyfncSYKuadoQm9pDlZ_3/view
In addition to those already mentioned Kimberly Brehm’s got a lecture series that I’m finding very useful.
https://youtube.com/playlist?list=PLl-gb0E4MII03hiCrZa7YqxUMEeEPmZqK
Thank you very much!
There's an MIT OCW course by Gilbert Strang which I liked. The plus is that he also has a book on the subject which you can refer to.
Thanks!
The YouTube channel 3blue1brown by Grant Sanderson has a series of videos called 'Essence of Linear Algebra' which is quite popular. Definitely worth checking out.
Thank you!
hello, last year I've learned some ML algorithms, and I've made around 5 projects.
and i just graduated from high school so I want to learn more about the
mathematics of Machine Learning before getting into university,
and I'll start in linear algebra, any recommendations or tips?
​
Mathematics for Machine Learning Course (Coursera)
this Course is good, I took half of the Linear algebra course,
but I want to grasp deeper knowledge
​
​
Mathematics for Machine Learning Book
this book is great, but I want to learn more and build a good foundation in these topics.
so most likely I won't go with it.
​
i have 2 choices for now
​
1:Linear algebra by Prof. Gilbert Strang
a lot of people recommended this course but I feel that it's too old.
​
this book is great and here's a full explanation of the book from the author.
​
i guess these are my choices and if you have any
tips or recommendations for Linear algebra, Calculus, and Probability & Statistics,
please write them down.
​
and what's your approach guys in learning math,
because I feel that it's too hard to use the spaced repetition technique in Mathematics.
I mean the learning process gonna be long, so I have to figure out the best way to learn
so I can grasp the Concepts and the topics well.
;)
I have not used these frameworks yet. But these look promising.
How to Teach Yourself Math by Scott Young A framework on how to find insights when learning maths.
Learn Difficult Concepts with the ADEPT Method A strategy used to help get an intuitive understanding of a maths topic.
Why do you feel Gilbert Strang's course is too old? The fundamental principles of linear algebra haven't exactly changed in about 100 years.
I do not know why I am not able to understand from him.
oh sorry, i didn't make it clear.
i mean the recording is too old.
I need some extra help trying to learn the concepts of linear algebra. I know of Khan Academy and 3Blue1Brown on youtube but I was wondering if anyone can suggest other websites or youtube channels to check out.
I taught Linear last year. These are the videos I made for my students. They are all extra practice problems and solutions.
https://youtube.com/playlist?list=PLscpLh9rN1Rfo0ifw9RZFoJ2Te2jk_pwX
https://hefferon.net/linearalgebra/index.html
chapters, exercises and solutions.
Try this one
Hey everyone, I am starting in Fall 2023. I am hoping to brush up on my linear algebra which I last studied some 20 years ago. I am not in a rush since I don’t need linear algebra for the first few courses I have planned. But rather I want to build a strong intuitive understanding, which is how I do math best. I tried doing the GT edX course, but I feel like it mostly deals with the mechanics of how to calculate stuff rather than intuitively understanding what it means. Some options that have been previously suggested on this sub Reddit are listed below - if any of you took these, could you comment on how you liked it?
Mit ocw. Gilbert Strang is very very good for linear algebra. Plus you have all the hws, exams and even problems specific to lessons to practice.
is the videos from 14 years ago still good? or are there specific playlist? https://youtube.com/playlist?list=PL49CF3715CB9EF31D
Yes. I’d suggest going to mit ocw website and finding the sc section of the course. You’ll find everything neatly organised there. Each lesson also mentions the related readings from the book, so if you don’t understand something you can always read the relevant sections from the book. This is tailored to the latest edition of the book if I’m not wrong, so you don’t have to worry about missing out on anything.
of course they are good, linear algebra is not something which is fast moving
3blue1brown has a good series on the intuition behind linear algebra.
I was going to recommend this. 3blue1brown does not provide the mechanical steps, but he helps you build the intuition to internalize the concepts very well
That’s the YouTube channel right? Ah thank you! I had stumbled on a really good video some time back, but couldn’t find it again. I think it was this one!!!
The Schaum’s book on Linear Algebra is relatively cheap, well written, and has many examples.
Notes from when I was a linear algebra instructor in a past life:
https://drive.google.com/drive/folders/1J7ndpMeg8Dx_XcV_Vg5ddGStHIi5neeM
I don't necessarily recommend them over the other resources you've mentioned, but I'm happy to share them.
Thanks all, I will get started with his lectures on MIT OCW (unless there is a different set of lectures?)
I believe he is the one teaching the MIT OCW course?
Hi everyone,
I’m looking to dive back into Linear Algebra, but I’m having a hard time finding the right book. I studied university-level math about 20 years ago, so while the foundation is there somewhere in the back of my mind, I definitely need a refresh, ideally something that’s rigorous but also explains the intuition clearly.
I’m not looking for a quick reference or just exercises, but a book that helps me understand and rebuild my thinking. I’d really appreciate recommendations that worked well for others in a similar situation.
Thanks a lot in advance! 😊
Your welcome. He has free videos on YouTube. A complete course in linear algebra from mit
Linear Algebra for Data Science by Cohen on OReilly press is working for me. Comes with a GitHub repo as it teaches both theory and code in python.
I have heard of this one, did not lnow about the github though, thanks!!
i used otto bretscher linear algebra for my class. had to take linear algebra after my bachelors to get into stat grad programs since i wasn't a math major
Thanks
I was in a similar spot, and “Linear Algebra Done Right” by Sheldon Axler really helped, it’s rigorous but explains why things work, not just how. If you’d like more intuition and geometry, Gilbert Strang’s “Introduction to Linear Algebra” is fantastic too. Both do a great job reconnecting the theory to the bigger picture rather than feeling like a list of techniques.
Thank you for your input!
Linear Algebra done right by Sheldon Axler is probably the best one. The treatment of most linear algebra texts is that from the perspective of matrices and operations on them, especially with the use of the determinant. In my opinion (and evidently the author’s opinion as well) this is a very unenlightening, algebraic way to look the subject. But if you approach the subject from the perspective of linear maps first, a lot of the subject matter becomes easier to understand (at the cost of more involved proofs). For example, the typical proof that there is always an eigenvalue for any linear map in any odd-dimensional vector space follows by considering the determinant of the matrix A-λI while Sheldon Axler approaches it by considering the minimum polynomial (the unique monic polynomial of smallest degree that annihilates the operator). I also liked that it introduces matrices as a consequence of linear maps, while most texts do it the other way around. This helps in my opinion with many questions like “why can’t we divide matrices” by making the analogy of “why can’t we divide by functions?”: because it makes no sense. When I say f(x)=y, do you think to divide by f? I guess it depends on what you want linear algebra for. If you want a theoretical knowledge of the subject, definitely go for this book. Otherwise, if you want a computational perspective, I’m sure any book works, Lay/Strang/Larson are the typical recommendations I believe.
You can "divide by" a matrix by multiplying with its (multiplicative) inverse, same as division of real numbers
You can "divide by" a function if you first define what it means to "multiply" a function
You can’t divide by a matrix. You’re just applying the inverse map. The analogy was meant to liken multiplying by the inverse to taking an inverse function. How do we solve f(x)=y? You take the inverse and get x=f^-1(y). (Given that the inverse exists). Now a linear map T on a vector space with a given basis is isomorphic to a matrix M(T). If the map is invertible, T^-1 is also a linear map such that T^-1 T=I (much like function composition). So since the inverse is also a linear map, it too is isomorphic to a matrix M(T^-1) (we take the matrix wrt the same basis as that of T). Then M(T)M(T^-1)=M(TT^-1)=M(I) is the identity matrix. This is how we “divide” by matrices. But we never divided by the matrix. We just used properties of linear maps and the isomorphism M from linear maps to matrices. The thing is people treat matrices as “just an generalisation of R” and that we can do everything with matrices as we can in R, but they’re kind of different things.
Clearly f(x)=y gives (x)=y/f after dividing by f
Thanks for your detailed feedback!
This is my own response to a question I've posted here: Original purpose
What I was trying to ask in that post was simply some online material to learn L.A as a mathematician (at least close). Keeping it short, most materials on internet, which sometimes are reflection of our school system, are kind of Wikipedia-ish. I've tried all kind of standard materials like Gilbert Strang's, David Poole, Sheldon Axler and Sergei Treil, but I wasn't getting it quite right.
So, I wanted something like this: Essence of Linear Algebra, by 3blue1Brown, which really showed the beauty of L.A; but I wanted something more complete. And I've found this amazing course: Essence of Linear Algebra, by MathTheBeautiful, which keeps the qualities of 3B1B, showing the power of creativity, deduction, pattern recognition and etc etc. Both materials are very different from the standard School/College approach. I'm watching this for 2-3 months now, but I thought I should post this for people in same position as I was.
P.S: I'm not them, nor I know them, nor I was paid.
why was axler not a good fit for you? i found it to be the best linear algebra resource i have come across anywhere
Sincerely, I'm not good at expressing myself to people. Most of the time people get me wrong. I'm not that good at social "things". Maybe someone else in future will see this post, and know what I'm trying to say. But maybe these materials (the standard ones) are better for people more interested into applications or how it works instead of why it works? idk
Before these coursers, I could perform matrix multiplication and calculate the determinant of 2-3D matrices. I could just do it without knowing why it works and how to come up with that. These courses make me feel like I'm discovering math and proving to myself, like the first mathematicians maybe. I find myself saying:
Like in Computer Science now, we are used to work with abstraction over abstraction. So for me, these standard materials are like that. We are using abstraction to get the optimal results, but without looking deeper on how and why it works. Now we use transistors, python, DvD and etc etc. But these are the evolution of some base. And by studying this "base" (low level maybe), you'll get a much more profound knowledge (even if it is more time consuming)
And after all this, I now find myself remembering less, but deducting more and more until I remember a formula, following so a path of reasoning. It's like I don't need anymore to remember F=ma, because I understand enough to come up with it after a moments of thinking.
FWIW math books are often in their problem sets as much, or even more so than in their main text (in fact, there are also wonderful problem books). You often won't get their beauty by reading them, but by working through them. I'm not implying anything about how you engaged with these linear algebra resources, just know that you could be judging fishes by their ability to fly.
Watching 3B1B since Calc 1, he's awesome. Will check out the other vids!
Also recommend Krista King, she's great too!
Has Krista king made linear algebra videos?
Edit: It looks like they are on her website behind a paywall. I may have to check them out
Yup! I’m subscribed right now and find her vids and walk through a well worth the sections I’m struggling with.
Thanks for the tip, I will
Just going through the video course by 3blue2Brown and it's really a blast. I plan on following up with MathTheBeautiful. Thanks for the recommendation!
I'm glad my post was helpful to you
Do YouTube courses include practice problems and tests?
Yes, but because there's no teacher to give it to you, you also have to search it yourself. And the more I do at least, the more I get better. It's like learning programming too. You start searching for projects and solving, until you move to a next subject. If it helps, here's the websites:
But this is internet. You're learning for free, but YOU need to do the work that your teachers would do for you. You're on your own. But there's pleasure on keeping a schedule and whatever
I'd like to really get the 'hang' of linear algebra so I'm confident in my spatial programming. I've used blender a lot and I seem to be comfortable with the concept of different types of vectors and spaces and using matrices to translate between them in my python scripts. Past that though, everything is very slippery.
I've cracked Lang and Axler, but I feel sorta over my head even in the first chapters. But the 3blue1brown videos are easy and tbh too simple. Surely there are some good resources 'in between'?
I would highly recommend this free ebook https://mathfor3dgameprogramming.com/
The chapters on linear algebra are very well done, in my experience. I hold a bachelor's in math and found lower division Linear Algebra quite challenging.
The book you’re reading can be a little intimidating if it’s your first go at linear algebra and have little to no experience with proofs. To add, it’s hard to appreciate the ‘done right’ aspect of the book without having taken a previous course that teaches the concepts in a different ordering. I’d recommend looking into any applied linear algebra resources as that would most likely give you a better foundation. Haven’t read it, but just glancing over, “Applied Linear Algebra” by Olver looks promising and about the same style as the course I took at my university.
I took an intro course that used Linear Algebra and Its Applications by Lay, Lay and McDonald and thought it was quite a straightforward book with a lot of mechanical exercises. There are .pdfs. I have only worked through the SVD section of Linear Algebra Done Right but it seems difficult to get much out of it without having taken intro level proof-based coursework. Strang has other texts as well that I believe are more traditional but his authoring style isn't for everyone. A lot of the content in Linear Algebra Done Right is not useful for programming in the graphics pipeline unless you are getting somewhat esoteric.
Thank you!
I followed this course:
https://www.youtube.com/playlist?list=PL_a9tY9IhJuPDEDq97tq0uKXpsTZYBIXe
It has a very wide approach to LA, but it is useful for more fundamental understanding of its base concepts.
I will check this out
Coding the Matrix is pretty sweet, an implementation-first tour of Linear Algebra which doesn't skimp on the actual mathematics.
Awesome, thank you!
best resources for learning linear algebra
Key Considerations for Learning Linear Algebra:
Textbooks:
Online Courses:
YouTube Channels:
Interactive Tools:
Practice Problems:
Recommendation: Start with Gilbert Strang's textbook or MIT OpenCourseWare for a solid foundation, then supplement your learning with visual resources like 3Blue1Brown and practice problems to solidify your understanding. Engaging with multiple formats (text, video, interactive) can enhance retention and comprehension.
Get more comprehensive results summarized by our most cutting edge AI model. Plus deep Youtube search.