it's an algorithm implementation model about H. Zou, T. Hastie, and R. Tibshirani. Sparse Principal Component Analysis in matlab.
I don't know how the two parameters work,I'm just setting it up.
Did someone has experience please explain it in detail
thank you!
ps：
Documentation comments
LAMBDA2, STOP
LAMBDA2 specifies the positive ridge (L2) term coefficient. If LAMBDA2
is set to infinity, soft thresholding is used to calculate the
components. This is appropriate when p>>n and results in a
significantly faster algorithm.
STOP is the stopping criterion. If STOP is negative, its absolute
(integer) value corresponds to the desired number of non-zero
variables. If STOP is positive, it corresponds to an upper bound on
the L1-norm of the BETA coefficients. STOP = 0 results in a regular
PCA. Supply either a single STOP value, or a vector of K STOP values,
one for each component.
and model file
I suggest you to follow: http://www.afarcome.altervista.org/srpca.pdf
https://people.eecs.berkeley.edu/~elghaoui/Pubs/SPCAhandbookSV.pdf
https://www.cc.gatech.edu/~hpark/papers/siamDM_sparsePCA.pdf
Follow this Link:
https://scikit-learn.org/stable/modules/generated/sklearn.decomposition.SparsePCA.html
- Badges
- Science topic
- Similar topics
- Comment