Timeline for Can we use MLE to estimate Neural Network weights?
Current License: CC BY-SA 4.0
14 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jan 29, 2023 at 19:14 | history | edited | Sycorax♦ | CC BY-SA 4.0 |
added 122 characters in body
|
| May 16, 2022 at 0:11 | history | edited | Sycorax♦ | CC BY-SA 4.0 |
added 277 characters in body
|
| Jul 29, 2021 at 22:16 | history | edited | Sycorax♦ | CC BY-SA 4.0 |
added 39 characters in body
|
| Jun 11, 2020 at 14:32 | history | edited | CommunityBot |
Commonmark migration
|
|
| Jul 25, 2019 at 17:58 | history | edited | Sycorax♦ | CC BY-SA 4.0 |
added 574 characters in body
|
| Jul 25, 2019 at 17:47 | history | edited | Sycorax♦ | CC BY-SA 4.0 |
added 141 characters in body
|
| Jun 18, 2018 at 2:55 | history | edited | Sycorax♦ | CC BY-SA 4.0 |
added 244 characters in body
|
| Apr 11, 2015 at 21:44 | history | edited | Sycorax♦ | CC BY-SA 3.0 |
added 269 characters in body
|
| Apr 11, 2015 at 20:10 | history | edited | Sycorax♦ | CC BY-SA 3.0 |
added 5 characters in body
|
| Apr 11, 2015 at 20:07 | vote | accept | tor | ||
| Apr 11, 2015 at 20:00 | history | edited | Sycorax♦ | CC BY-SA 3.0 |
added 78 characters in body
|
| Apr 11, 2015 at 19:38 | comment | added | bayerj | I beg to differ with what you say. The different local minima arising from symmetries are all of the same quality, so you don't have to worry about that at all. What you probably want to say is that ANNs do not have convex loss functions, which makes optimisation more involved and does not guarantee finding a global optimum. However, there has been quite some evidence recently that ANNs actually do not have that much of local minima issues, but rather saddle point issues. See e.g. arxiv.org/abs/1412.6544. | |
| Apr 11, 2015 at 19:32 | history | edited | Sycorax♦ | CC BY-SA 3.0 |
added 360 characters in body
|
| Apr 11, 2015 at 19:25 | history | answered | Sycorax♦ | CC BY-SA 3.0 |