Timeline for Nearest matrix orthogonally similar to a given matrix
Current License: CC BY-SA 3.0
26 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jan 11, 2019 at 20:34 | comment | added | Wuchen | @Suvrit Could you please explain why Hermitian matrices are easy in this problem? Thank you! I'm assuming that you mean A and B are Hermitian. | |
| Apr 25, 2018 at 2:35 | comment | added | fibonatic | Have you considered writing it with Kronecker products: $(I \otimes A - B^\top \otimes I) \mathrm{vec}(T) \approx 0$. When taking the svd of matrix multiplied by the vectorized $T$, then the right-singular vectors corresponding to the smallest singular values might be good initial guess for $T$. These will probably not be correspond to orthogonal matrices, so for each reshaped vector you could calculate the nearest orthogonal matrix again using svd. | |
| Mar 10, 2018 at 21:14 | history | edited | Turbo | CC BY-SA 3.0 |
added 98 characters in body
|
| Mar 10, 2018 at 21:10 | comment | added | Turbo | Oh I see ok.... it is trivial for those cases :) | |
| Mar 10, 2018 at 21:07 | history | edited | Turbo | CC BY-SA 3.0 |
added 98 characters in body
|
| Mar 10, 2018 at 20:58 | comment | added | Turbo | @YemonChoi yes.. | |
| Mar 10, 2018 at 16:57 | comment | added | Yemon Choi | Turbo, do you know how to diagonalize hermitian matrices? | |
| Mar 10, 2018 at 16:41 | answer | added | Rodrigo de Azevedo | timeline score: 2 | |
| Mar 10, 2018 at 0:31 | answer | added | Mark L. Stone | timeline score: 2 | |
| Mar 8, 2018 at 14:59 | answer | added | Bill Bradley | timeline score: 3 | |
| S Mar 8, 2018 at 12:12 | history | edited | Turbo | CC BY-SA 3.0 |
Added tag.
|
| S Mar 8, 2018 at 12:12 | history | suggested | Rodrigo de Azevedo |
Added tag.
|
|
| Mar 8, 2018 at 10:00 | review | Suggested edits | |||
| S Mar 8, 2018 at 12:12 | |||||
| Mar 8, 2018 at 0:18 | comment | added | Turbo | @Suvrit why so? Hermitian is an important case. | |
| Mar 8, 2018 at 0:17 | comment | added | Suvrit | if the matrices are hermitian then the answer is easy, not yet sure about the general case | |
| Mar 7, 2018 at 21:28 | history | edited | Federico Poloni | CC BY-SA 3.0 |
better title, tags
|
| Mar 7, 2018 at 20:54 | comment | added | Turbo | Either Frobenius norm or spectral norm is ok. | |
| Mar 7, 2018 at 20:54 | history | edited | Turbo | CC BY-SA 3.0 |
edited body
|
| Mar 7, 2018 at 19:33 | comment | added | Dirk | Could you please clarify what you mean by $\|\cdot\|_2$? | |
| Mar 7, 2018 at 15:56 | answer | added | Igor Rivin | timeline score: 4 | |
| Mar 7, 2018 at 14:30 | answer | added | Dirk | timeline score: 6 | |
| S Mar 7, 2018 at 13:41 | history | suggested | Amir Sagiv |
relevant subject tag
|
|
| Mar 7, 2018 at 11:44 | review | Suggested edits | |||
| S Mar 7, 2018 at 13:41 | |||||
| Mar 7, 2018 at 11:10 | review | Low quality posts | |||
| Mar 7, 2018 at 11:30 | |||||
| Mar 7, 2018 at 11:05 | history | edited | Turbo | CC BY-SA 3.0 |
deleted 47 characters in body; edited title
|
| Mar 7, 2018 at 10:49 | history | asked | Turbo | CC BY-SA 3.0 |