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 !!!
bahaha la même j'ai un exam demain ça m'a mis bien
Dude, many thanks. There are not a lot videos on SVD in RUclips with that simple go-through example.
You just answered my question perfectly, this was very helpful. Thank you so much!
Thank you for taking the time to coment.
thank you very much!!!! I was almost crying because I couldn't understand this and you made it seem so simple.
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!
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!
Thank you so much, clear and concise!
This is exactly what I was looking for. Thank you so much
Glad it was helpful!
Thank you so much, you did not confuse me like my so called "professor" did
Great way to teach the concept of SVD. Thanks
This is the best explanation to determine SVD so far in RUclips
Thank you!
very true
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.
Excellent! Thanks
I thank you so much
Thank you!
You're welcome!
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?
thanks
10:25 Sir, could you give us a more specific clue that we might end up getting "oposite unitvectors"?
thank you so much, I have an exam today
You can do it!
Thx bb
All that effort to get to the same Matrix as a result .......
What happens if an eigenvalue is negative?
Niubi，bro
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!