[R] Partial Least Squares
andy_liaw at merck.com
Sat Aug 21 04:21:40 CEST 2004
> From: Spencer Graves
> Did you "read the posting guide!
> http://www.R-project.org/posting-guide.html"? In particular, did you
> try www.r-project.org -> search -> "R site search" for "partial least
> squares"? I just got 65 hits for "partial least squares". Some were
> not relevant, but many were. I also got 9 hits for "nipals", all of
> which seemed relevant.
> Also, have you considered "sem" (structural equation
> From what I've heard, partial least squares started out as a
> without a clear problem statement, i.e., an algorithm
> claiming to solve
> a problem but without a clear statement of a probability
> model for which
> their algorithm produced the (approximate) maximum likelihood or
> Bayesian posterior mode. Structural equation modeling, by contrast,
> provides the model and problem statement that partial least squares
> seems to try to solve. I got this impression from reading the PLS
> article in the Encyclopedia of Statistical Sciences.
AFAIU the problem statement for PLS is not that different than that of
canonical correlations. In fact some had criticized PLS as re-invention of
the canonical correlations. It can also be seen as a regularized (thus
biased) regression fitting method.
As has been pointed out, the pls.pcr package on CRAN offers PLS (both De
Jong's SIMPLS and the `kernel PLS' algorithms). There's also the gpls
package that combines PLS and the IWLS iterations in glm.
> hope this helps. spencer graves
> Lana Schaffer wrote:
> >Is there a Partial Least Squares package implemented in R?
> > [[alternative HTML version deleted]]
> >R-help at stat.math.ethz.ch mailing list
> >PLEASE do read the posting guide!
> R-help at stat.math.ethz.ch mailing list
> PLEASE do read the posting guide!
More information about the R-help