0

Given a dataset PCA can be performed via 3 ways:

  1. Eigenvalue decomposition
  2. Singular value decomposition
  3. Non-linear iterative partial least-squares algorithm

Can anyone shed light on comparative study of the 3 ways, what is pros and cons of each way? when to prefer a particular method?

kjetil b halvorsen
  • 63,378
  • 26
  • 142
  • 467
user13744439
  • 31
  • 2
  • There actually has been answered questions like this one. Though they don't necessarily refer to a "comparative study". Here is one https://stats.stackexchange.com/q/79043/3277. Search other yourself, please. – ttnphns Nov 05 '21 at 08:36
  • @ttnphns Correct. But, I don't see any question on Non-linear iterative partial least-squares algorithm for PCA and its comparison with say Eigen decomposition. Please mark this as duplicate with link to any such question which discusses Non-linear iterative partial least-squares algorithm vs eigen decomposition. Thanks. – user13744439 Nov 05 '21 at 13:38

0 Answers0