[MRG+1] DOC manifold examples added to docstrings #11823
Conversation
|
What do you mean? You mean more examples? I'm not sure what "this" refers to :P |
|
I mean I can't immediately think of a better way to simply illustrate
manifold learners, but I'm not a big user of them.
|
|
Ah yeah, in our department we used it simply as a nice dimensionality reduction technique, with the extra sugar of actually plotting the embedding, then coloring the clusters/classes to somewhat "confirm" the validity of the embedding. Otherwise we used it like these examples. |
|
Is it meaningful to apply manifold learning to Friedman #1 dataset? |
|
I actually started with the digits dataset, but the tests were too slow. I kinda preferred a generated dataset just to have the sample code cleaner than taking a small [randomly] chosen part of the digits dataset. Also as long as dimensionality reduction goes, I don't see why friedman would be a bad choice. Still, I can change the examples if you really think it should work on a different dataset. |
Maybe we can use something like
Yes we can reduce the dimension of everything, but according to the definition of friedman, I can't understand what you'll get after dimensionality reduction (and how you benefit from dimensionality reduction). |
See #3846
Examples added to classes under
sklearn/manifold