gfhf

Good

nice

nice

The stated goal of the course is to provide a sufficient base of knowledge in linear algebra for applied data science i.e. (a) to teach linear algebra without gory proofs or endless grinding through algorithms by hand and (b) to foreground geometric interpretations of linear algebra that can be recalled for many data science techniques and visualized with common data science tools. While I appreciate this goal and enjoyed the early foray into projection, I never felt the "a ha" moments I did as an undergrad in a class that used Gil Strang's "Introduction to Linear Algebra" (which I reread alongside this course as a supplement). The course seems to ask for some faith that various concepts introduced earlier in the course will be united by the end, but never makes good; opting instead for a kind of sleight of hand: having students implement the Page Rank algorithm with the intention that this will draw together the core concepts of the course. It could be that I was just looking for a more complete treatment of the subject than the course ever intends to offer, but I strongly felt that with a bit of restructuring, that the subject could be presented primarily intuitively, but with a level of clarity and artfulness in its conclusion that will ensure that students remember the core concepts beyond when they remember its presentation.

As of this writing, I am almost done with week 4 of Mathematics for Machine Learning: Linear Algebra. The content of the course is excellent and professor David Dye's lectures are to be commended no doubt. The reason for my low rating is because the programming assignments are broken and that's really not acceptable for paid offering such as this. To clarify, at various points throughout this course, students are asked to complete a programming assignment. The student is presented with a button which says, "Open Notebook". The student is supposed to click this button and be redirected to a Jupyter Notebook (and interactive Python execution environment). Unfortunately, instead of being redirected, click on this button results in a "404 Not Found" error. There are various discussions in the class discussion forum regarding this issue (some months old), but no action has been taken to resolve this issue. Luckily, someone taking the course managed to find the programming assignments and posted them on google docs for others to use. I've been working these which is fine, but as I said, we're paying for these courses, someone should be resolving this.

I never took Linear Algebra in university. The last time I got exposed to this topic was more than 10 years ago when I was still in junior high. This course is very condensed. Each video covers each topic relevant to ML very briefly and the instructors go very fast on explaining each topic. This means students have to do a lot more research on their own to really comprehend the concepts. What's nice about this course is the programming assignments. They give you a chance to apply math concepts to the computational model. Something like this you wouldn't have a chance to do if you don't spend on an online course like this one, I guess. Overall, I think this course provides values in a way that gives you an overview of how Linear Algebra is used in ML. For me personally, I know I still need to consult other sources online to further understand Linear Algebra as I'm not sure that after finishing this course I've got adequate knowledge to pursue ML. What all that said, hence I give this course 3 stars.

I personally do not recommend this course for anyone who has not had any prior knowledge in linear algebra. Although it states that this course is for beginner, you should have a (quite) decent background in linear algebra to understand all the materials. The geometric intuition parts of projection and eigenproblem are interesting. I would suggest the instructors to prepare some lecture notes for each week, instead of giving just a single cheatsheet at the end of the course. This would help learners to review the material better. Besides, some videos do not have the subtitle synced. Thus, matching between the full transcript and the video is quite hard for non-English speaker.

The lecture material in this course is great, and the quizzes are a lot of fun and it provides good resources for learning. However, the programming assignments are a pain due to lack of guidance and the grades are penalized due to minor things like indentations as opposed to actual math errors. This isn't a python course, its a math course, and grades should be awarded and penalized based on the math skills one has acquired throughout the course, not on the programming or whether an indentation is off. I highly recommend the course to learn linear algebra but I strongly encourage the instructors to improve the programming assignments or alter the assessment methods.

It must be difficult to pitch the level of these courses.

I have been taught Data Science whilst on an apprenticeship but didn't feel the maths was taught rigorously enough and hoped this would fill gaps of in knowledge.

The breadth of the concepts covered on the course achieved that but a lot of research was required from other resources to clarify certain topics which is why I think a beginner rating for this course might not be fair.

If you are not confident with maths, this course is achievable but expect to devote time to on other sites.

The PDF supplement is concise but useful for reference

T​his course was helpful in reviewing topics in linear algebra and focusing on the usefullness of projections and eigen problems in manipulating vectors. If you have taken a linear algebra course previously you should find the course pretty easy. If you have not taken a linear algebra course before, this course may be much more difficult to complete unless you are prepared to use other resources to help manage the course topics and assignments. I was hoping for more challenging course assignments but understand that much of it is self-assessing what you are learning as you go along.

Sometime the course does not clarify some principle. Also, I still cannot understanding that why the eigenvectors have relationships with page rank and why can we use the probability of reaching the link to each page as a vector. I cannot understand the relationship. Plus, the final quiz contains something that I have not learnt in the course, such as damping. I still cannot understand the Quiz2-5. I learn much in courses week 1-4, but I am much confused about the week 5. Thank you for listening.

Fiirstly, going with the positives , the instructors were clear and effective in teaching the subject. Also,the feedback from the assignments were also good .Video quality was amazing.

I also felt that it was a very brief course, not worth an average Indian father's one week income.Also there was no option for Audit. Also, most assignment were substandard and involved lot of calculations which I felt is a waste of time. The coding assignments were also pretty simple and straight-forward.

Your discussion forum really needs to be better. It seems to be the only place where any sort of doubt clearing can be done and very rarely have I seen TA's answering unless it's a grading issue. The problem with this sort of answering is that if any coding concepts are unclear, either they are solved by trial and error or after going through Google multiple times. And for a course that is paid for, I shouldn't have to make this much of an effort just to have my doubts cleared.

The course starts well and in general the first instructor does a good job trying to help the student develop an intuition of the concepts. However, weeks 4 and 5 are extremely weak. Very important concepts like eigenvalues and eigenvectors are poorly explained. The final quiz on these concepts asks questions that were never discussed or explained in the videos. I found I needed to go elsewhere on the Internet (like 3Brown1Blue) just to help me get through some of the quizes.

Teaching quality is good overall, except for a few jumps towards the end, where it's hard to follow. Quizzes and assignments well designed.

Unfortunately, and contrary to other courses I've taken, the forum seems completely un-monitored (as of May 2019), so don't expect much help from there.

Overall it's a good start of the specialization. Sadly, the teaching quality of the other two courses (multivariate calculus and PCA) is way below the standards of this one.

Without about intermediate knowledge on vectors, you would be having a tad difficult time doing this course. I have already studied Linear Algebra in my University, attaining an A grade and found the explanation for some of the parts a bit confusing. It is not bad by any means, but just that I believe a more structured/concise explanation for some of the terms might be more accessible for beginners or students who have never learned about Lin Alg before.

The course was informative on most of the Linear Algebra you need for machine learning. The programming assignments and quizzes are mostly relevant to what you learn beforehand. However, some of the videos feel rushed and I found it difficult to take notes in time without having to replay the videos. There is also a lack of written content after each video describing the content covered, which would have been a godsend for me with my method of learning.

Definitely enjoyed some parts of the course but in general, the explanations are brief, requires spending significant time outside of videos on Youtube, discussion boards etc as they skip or miss key points for a beginner to grasp Linear Algebra concepts. Happy that I completed the course, but a lot of improvements should be made by including course notes that supplement common queries/misunderstandings students have in relation to the course materials.

This course is not very hard for students with some maths foundations like me, but the programming assignment is too hard, even though I knew some basic Python knowledge. Two lecturers sometimes could not explain one concept clearly with some typical examples. I could only learn by doing assignments or use formulas to calculate without real understanding. This course is only for some basic concepts but not solid learning.

The videos were well structured, but the quiz sometimes were far more difficult than the practice questions in video. I had posted on forum but no comment nor reply. Quiz answers were not elaborate enough to understand after making mistakes. So I had to ask my father who's extremely good at maths many times, for explanations. Without hi help I might not have been able to understand as well. Need improvement.

This class fluctuates between impossibly hard, because a lack of instruction and examples were provided and too simple, because the same question is asked repeatedly. There is also very little mention of machine learning, despite the name of the specialization/course. That said, it was still the best survey course of the linear algebra concepts most relevant to machine learning.

The course starts off well enough, but by week 4 the intuition for certain concepts is not imparted well at all, and the correspondence between notation from the lectures and that used in the practice quizzes breaks down badly.

I gave it 3 stars instead of 2 stars because the geometric intuition that is imparted is quite good, even though at times the notation is sloppy!

Having no background in linear Algebra made it difficult to complete the quizzes, assignments and exams. Even with the instruction (which was good) I found the hands on portions to be different from what was being explained in the videos. I will instead have to take the key concepts and do more research on my own to fully understand them.

Let me first thank you all for the effeors that you put to make this course avaialbe on Coursera. The teachers are passionate and they have their unique ways of sharing their knowledge. However, I think some warp ups are needed after each cocept as at some point i got confused and did not know what we were doing this and that.