Determining the eigenvalues is significantly faster when you do AA^T instead of A^TA - anytime you have a wide matrix you can compute it faster (i.e., if you have a 2x5 matrix you have to do the eigenvalues of a 2x2 instead of a 5x5...) & vice versa for tall matrices. Because of eigenvalue properties e-vals of A^T = evals of A in this example since it's symmetric, useful tip if you want to speed up your computation!
Since V is 3x3 and U is 2x2, in this case could we find the SVD of A transpose, then transpose the resulting matrices to get the SVD of A, so we can work with a 2x2 instead of 3x3 matrix and save some time?
10mn explication in English easier than hours with my teacher in France thank you !!!
thank you very much!!!! I was almost crying because I couldn't understand this and you made it seem so simple.
You just answered my question perfectly, this was very helpful. Thank you so much!
Thank you for taking the time to coment.
Dude, many thanks. There are not a lot videos on SVD in TH-cam with that simple go-through example.
Straight to the point, before this video i watched 4 other videos, jesus christ, thank you very much
Thank you for your comment. I hope it helps the video rise in the search algorithm. 😂
Just explained in 10 minutes, what my prof failed to explain in 2 hours
Glad I could help!
This is the best explanation to determine SVD so far in TH-cam
Thank you!
very true
Thank you so much, clear and concise!
Determining the eigenvalues is significantly faster when you do AA^T instead of A^TA - anytime you have a wide matrix you can compute it faster (i.e., if you have a 2x5 matrix you have to do the eigenvalues of a 2x2 instead of a 5x5...) & vice versa for tall matrices. Because of eigenvalue properties e-vals of A^T = evals of A in this example since it's symmetric, useful tip if you want to speed up your computation!
so simply explained and covered all important nuances to remember. thank you
Thank you!
Bro clutched on my linear exam today
W about to take mine aswell
Superb. Very clear and helpful. Thanks for making the effort!
Thank you. You are very welcome.
Great way to teach the concept of SVD. Thanks
Excellent! Thanks
10:25 Sir, could you give us a more specific clue that we might end up getting "oposite unitvectors"?
Since V is 3x3 and U is 2x2, in this case could we find the SVD of A transpose, then transpose the resulting matrices to get the SVD of A, so we can work with a 2x2 instead of 3x3 matrix and save some time?
thank you so much, I have an exam today
You can do it!
I thank you so much
Thank you!
You're welcome!
thanks
Thx bb
What happens if an eigenvalue is negative?
Niubi，bro
White Melissa Lewis Dorothy Thomas Joseph
why is the sigma matrix sum a 2x3 matrix
Sigma will always be an m by n matrix so the multiplication is possible.
Thank you!