header image
Home arrow Tutorial arrow A statistican's view of PLS
A statistican's view of PLS PDF Print E-mail
Written by Dick Kleinknecht,Weyerhaeuser Company   
A problem many statisticians have with chemometricians is that we don't always understand what you are talking about. I'm sure chemometricians have a similar problem with statisticians - even though we are both deeply interested and involved in the same pursuits.

We both use similar words to describe our work, but I suspect the same words mean different things to chemometricians than to statisticians and vice versa. Most people in either discipline are too nice to mention this. The world's scientific literature is full of PLS tutorials by chemometricians. There are far fewer written by statisticians who are also regular PLS users and converts to that technology. I am such a person and will describe the NIPALS algorithm for the single-Y problem as a statistician understands it. As the description unfolds, you will see what I mean about our separate uses of some terms.

Download this tutorial (pdf)
< Prev   Next >
Search website

Editorial flash
Svante Wold's New Year Editorial 2005
Another year has passed too quickly, and we now read January 2005 on our calendars. In 2004, good things happened and terrible, happy and sad, as usual. However, chemometrics does well, spreading deeper into biology where the data sets are larger than ever with thousands of variables in, among others, gene arrays, LC-MS profiles, and NIR and IR microscopy.
Read more...
News flash
Chemometrics from an industrial perspective'

The Belgian Chemometrics Society (BCS) kindly invites you to 'Chemometrics from an industrial perspective', a BCS session integrated in CAC2010 with the presentation of the 3rd D.L. Massart Award in Chemometrics'. Though integrated in CAC2010, people can apply to the BCS session only through the BCS website.

Kind regards,
The BCS committee
http://chemometrie.kvcv.be

The D.L Massart Award has been made possible by the financial support of:

CAMO
CQ Consultancy
ExxonMobil Chemical
Infometrix
Johnson & Johnson Pharmaceutical Research & Development
Minitab
Novartis Pharma
Shell
Solvay
Vrije Universiteit Brussel

 

 

 
Tutorial flash
Multivariate Solutions to Metabonomic Profiling and Functional Genomics, Part 2
Simple, unsupervised chemometric methods obviously work well for data sets with a limited number of well-defined classes. However, biological systems are seldom simple and many of the biofluid datasets generated within metabonomics require more sophisticated statistical data analysis.
Read more...