The course is very good, almost perfect for my purposes. I liked specially the effort to make the students get the necessary intuition instead of pushing a lot examples as many other MOOC usually do. But I've noticed some negative points. I ask you to take my critic as a sincere effort to improve the course and eliminate some mistakes that really matters to the students. The last quiz seems quite disconnected with the lectures and there isn't a support guide or tutorial not even a mentor answering the questions in the week 5 forum. Some mistakes on videos (eigenvalues and eigenvectors) were confirmed by the lecturer but never corrected. Not even a errata on resources section. Talking about the resources, I think it is very poor. Cousera has many better examples.

The content and the speed are not satisfactory.

The speed totally hampers the content, lots of things aren't explained especially after Sam took over in the last module.

Other than the first 2-3 intuition videos and the programming assignment nothing was good in the 5th module/week.

It was very very difficult to follow the page rank video. I still don't understand it. For eigen basis I had to refer to other material outside this course.

This course is excellent however it is not for the mathematically immature unless they are willing to put quite a bit of additional work in. Arguably it can be classed as "Beginners" but still, I can imagine many will feel lost very quickly. At one stage David Dye offhandedly mentions soh-cah-toa... and that really sums up a lot of what is required in terms of mathematical maturity - high school maths at a reasonable level.

Those that undertake the course should be assisted by referring to additional materials when they feel things are a bit of a struggle, I did, and this greatly helped, although my Maths was around UK high school level (in Algebra and Trig).

Overall first class and easily manageable with a little work!

the instrutors were too good and their explination for the concepts was to the point and it made me realize things in linear algebra I didn't know before although I studied it in school of engineering

This course is phenomenal, It helped me to refresh a lot of skills that I learned at my college and at the same time I learned a bit on how to introduce all this matrixes into a programming assignment which are by the way extremely hard because I am a novice at programming. It helped me to see other subitems such as Gramm Schmitt and eigenvectors that I did not see on college, I understood them but not a 100%, I guess an 75% is an average. Thanks Coursera and Imperial College London for this awesome course. I had to search other books to comprehend the subject, but next time, be more detailed.

I only completed three out of the five weeks of this course. Too many of the lessons were just a source of frustration for me. The instructor doesn't explain things very well. For example, with change in vector basis, he walked us through using the dot product and scalar values, but then added them up. Nowhere did he say the last part was just a check, and it had me confused for quite a long time. Then, with Einstein's Summation Convention, he doesn't really explain the subscripts and what rules there are for their use. Plus, it's hard to follow along because he says the math out loud, then just writes down the answer. Since this is new to me, it would be good to see it written out, like | (1/2)(-2) + (-1)(4) |. Far too often, I had to rely on other resources to get enough of an understanding to complete the quizzes. By the fourth week, I started just skipping to the quiz and finding other resources to teach me how to solve the problems. Then, I decided to just give up entirely. And finally, there were issues with the auto-grader. With one, I needed to write out the values as 2.0 instead of 2, but there was no mention of needing this precision. With another, it was A[3, 0] (with a space) instead of A[3,0] (without a space), even though the provided code used A[1,0] (without a space).

Course is titled incorrectly. The course has nothing to do with machine learning. It's mainly out of context symbol pushing (like most math courses).

I expect any positive reviews will be from folks who do not work as a practitioner in the field and just want to promote "good vibes". Beware if you're actually looking for contextualized understanding, as this is not the course for you (at least through the end of week 3).

This is indeed one of the best math courses I have ever done in my life. This course changed the view I look at matrices and vectors. I have been 'transformed'.

The instructors were simply amazing. Totally loved every bit of the course. Amazing way to teach this math course, with proper motivation and intuition.

And for all the people writing negative reviews about no Machine Learning being taught in this course, it is clearly mentioned that this course teaches the math which is required for learning Machine Learning and not Machine Learning itself.

2 stars because this is not a beginner's course. false advertising. I'd recommend checking out some Khan academy videos on linear algebra and have at least a beginner's understanding of programming before attempting this course.

Not intuitive or well explained. Examples are horrible

1. I had to go through youtube videos for almost every lesson in this course to understand what is going on! For example, the video preceding "Practice Quiz: Linear dependency of a set of vectors" only explains the theory and the practice quiz is all about using the theory to solve. From my perspective, I just watched someone who knows math and is flaunting his knowledge instead of actually showing how to approach a question. The questions provided in the quiz are there without a proper explanation or an example on how to solve them. 2. Whenever a theory/solution is explained, the lecturer doesn't pause or emphasis that this is a point. Instead, he goes on as quickly as humanly possible. Additionally, the drawings/board writing illustrations end up like a bowel of spaghetti after he is over, good luck figuring out what happened there! 3. Not enough practice questions, and even for the questions that are there, there isn't a clear guide/help on how they should be approached in case a student got stuck.

First of all, the instructor clearly loves the subject he is teaching. You can tell immediately by the voice and the gestures.

Second, the fact that he is not a pure mathematician means he is constantly looking for the link between what he is teaching and practical examples. That's a must when you are teaching math to students intending to use it in real life (Machine Learning).

Third, there is a good structure to the material being taught, always building on what has previously been taught.

Fourth, is the amount of quizzes and exercises. Math can only be learned effectively if you keep challenging yourself in quantity and quality. Everyone remembers the quality bit, but some miss the quantity. Not this instructor I have to say. Congrats for that.

Fifth, intuition is being built from day one. Big applause for that as Linear Algebra lives and dies by the amount of intuition that's being put into its practice.

Sixth, my hat's off for the esthetic quality of the figures and exercises, and for their clarity.

Seventh, even for an engineer like me (graduated over 20 years ago I have to say) who's been best friends with Linear Algebra and Calculus for the 6 years I was in university (10 semesters + a final Master's thesis), this course wasn't trivial, and kept me making an effort at every bend, at every corner. This is something I am grateful for, as while I was refreshing concepts that I hadn't touched for 20 years now, I did have real fun.

Eighth, the coding examples are a magnificent tool that greatly helped strengthening some concepts (like Gram-Schmidth, etc.). Amazing job there.

Ninth, you can see how towards the end of the course, gears are shifted, pushing the student to get mental agility, and conveying the student the importance of intuition building (+ some algebraic symbolic manipulation) as opposite of focusing mainly on the symbolic manipulation. A good approach nowadays that computers do the computation for us, as opposite to what it used to be some decades ago. I really liked the fact that the instructors (Dr. Dye and Dr. Cooper) tried, both, to covey this very practical philosophical paradigm from day one.

My most sincere "cum laude" score for this course.

Nabil Chouaib.

Pros:

Amazing explanations of the covered topics, extremely engaging teaching staff

Focusses on the right things

Good and enough practice problems

Great (albeit easy) programming problems

Cons:

Calculation of Eigenvectors could have been covered better in my opinion

A final handout for all the covered topics would be really nice

Overall a tremendous course if you want to brush up on linear algebra. To me LA was taught mostly doing rote calculations without motivating the concepts or explaining them geometrically. I had more than a handful of "oh, so that's how this actually works" moments. I feel like my intuitive understanding of linear algebra concepts has made a big improvement.

I took a great pleasure to study this linear algebra course, teachers are very talented since their way to explain mathematical concepts make it very easy to understand , in fact with this particular amazing approach I changed my perception about learning math and sciences in general. I do recommend this course if you look for a global overview of linear algebra for direct application in machine learning or computer sciences!

A very good introduction but some of important content need to use another provider (Kahn academy) to understand completly

The instructors are good at teaching, but they don't teach you enough. so brace yourselves for long hours of looking at the screen, suffering from imposter syndrome. i truly am scared of taking the next courses.

Lots of unaddressed inconsistencies.

Mainly explains how to operate with matrices and vectors. Not how to use those in machine learning. If you expect to have a clear view of the usefulness of eigenvectors and eigenvalues in machine learning, this is not your course.

This course is seriously lacking. I spent almost all of my time on Khan Academy due to the horribly high-level video lectures. The lectures in this course threw out strategies and ideas without explanation of the foundational mathematics behind them. Then the students are expected to apply these unexplained strategies to complex problems in the quizzes.

On the other hand, the programming assignments were a joke. They provided zero learning experience as everything needed to solve the assignments was provided in the already completed sections of code. Minimal thought was involved. This course needs a serious overhaul. These are complex topics that require more than a 4-10 minute video to explain. I understand that this is not an actual college course, and as such the professors cannot go as in depth comparatively. However, in its current state, this course has given me little understanding of linear algebra. I would suggest potentially increasing the price of the course so more in-depth lectures can be provided.

First code lab (week 3- identifying special matrices) is broken- as I can see in discussion forum learners had problems with opening it many years ago. There are many topics about this problem in discussion forum, few answers from teachers or mentors (and no answer from any technical staff who are probably responsible for that error) and no solution for this problem.

No help at all - instructors have abandoned the course. I emailed Prof Dye with a question but got ignored. I had to go to Youtube to watch 3Blue1Brown's channel on Linear Algebra and also ask questions on reddit/learnmath. The final assignment had such intricate calculations, that I just kept guessing and submitting. I do not recommend the course.

There has been no reading material for this course. This has not helped me, it is a big fault with the course.

I notice the option to unenrol from this course is not available. This is another problem, I have to quit this MOOC.

no Use

I feel like this course is underrated for people who want to learn machine learning. This, coming from someone who never did engineering degree. From the reviews seem like people were not satisfied with the lectures, but since week 1 they recommended plenty of Linear Algebra textbooks and Youtube. I like it because it encourages self-learning than being spoon-fed by the lectures. Not going to lie, this was the most challenging coursera course I have taken so far but that means I actually spent hours studying! My tips would be to watch the Youtube videos they recommend early in the course and attempt all the ungraded exercises. Utilize the discussion forum if you are stuck. I find the discussion forum and Youtube playlist really helped me grasp the concept. If you can get the textbooks, it's not necessary but they are also great study supplements

Amazing course!

Look, two recommendations about this course: this is a tough course! Especially if you've never seen Linear Algebra. Don't let this course be your first contact with Linear Algebra. If you do, at least take a famous reference book like Strang and follow the course with the book. Also, do not expect to watch the videos and understand the content magically. The course is full of assignments and challenging exercises, which are fundamental to having fun with the MOOC. This course helped me to review knowledge and, more than that, to develop new intuition about some topics in Linear Algebra. The instructors are great! The video production is really good! The pace, the assignments, the connection with Numpy and Python. I have only good things to say about this MOOC.