daviddoji/18065
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

lecture 28 notes

Browse Source
This commit is contained in:
Steven G. Johnson 2023-04-19 14:12:34 -04:00
parent e6c2b608f1
commit 31c9a11ebc
1 changed files with 6 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
README.md
View File

@ -322,3 +322,9 @@ http://dx.doi.org/10.1137/S1052623499362822) — I used the "linear and separabl
* Backpropagation for neural networks.
**Further reading:** Strang section VII.3 and [OCW lecture 27](https://ocw.mit.edu/courses/18-065-matrix-methods-in-data-analysis-signal-processing-and-machine-learning-spring-2018/resources/lecture-27-backpropagation-find-partial-derivatives/). You can find many, many articles online about [backpropagation](https://en.wikipedia.org/wiki/Backpropagation) in neural networks. For generalizing gradients to scalar-valued functions of matrices and other abstract vector spaces, what we need is an inner product; we covered this in more detail in [lecture 4 of *Matrix Calculus* (IAP 2023)](https://github.com/mitmath/matrixcalc#lecture-4-jan-25). Backpropagation for neural networks is closely related to backpropagation/adjoint methods [for recurrence relations (course notes)](https://math.mit.edu/~stevenj/18.336/recurrence2.pdf), and [on computational graphs (blog post)](https://colah.github.io/posts/2015-08-Backprop/); see also [lecture 8 of *Matrix Calculus* (IAP 2023)](https://github.com/mitmath/matrixcalc#lecture-8-feb-3).
## Lecture 28 (Apr 19)
* Non-negative matrix factorization — guest lecture by [Prof. Ankur Moitra](https://people.csail.mit.edu/moitra/).
**Further reading:** Coming soon.