This is a well-designed course which achieves the declared intention of demonstrating the potential of PCA as a display technique and providing training in the use of one of the available programs. "
Journal of Chemometrics

Additional copies of the manual may be ordered separately.

This attractively packaged set, comprising 4 videos, a manual and a software package, is an introduction to the use of principal components analysis (PCA) and related methods in chemometrics. Emphasis has been placed on the use of PCA to display graphically the structure of data sets or to extract graphically information from such a set (display methods). However, links are provided to several other important methods, such as evolving factor analysis, principal component regression and partial least squares.
In order to induce students to learn PCA, and convince them of the usefulness of the methods, several real-life examples have been included. These examples have been chosen to illustrate the generality of the data analysis approach. The data pertain to food, industrial and environmental analysis, animal experimentation, medicinal chemistry, virology and epidemiology. In the software section an additional example concerned with food analysis has been added. Section 3 of the manual contains a complete list of figures. This allows the user to look at some of the figures in a more leisurely fashion or to re-read text which has been heard when viewing the videos. Some of the visuals and some of the texts (occasionally shortened) have been included in this section.
Two types of software augment this series. The first is a tutorial version of a commercial software package called SPECTRAMAP, which has been modified for didactical use. SPECTRAMAP is a performant software for methods derived from PCA that gives excellent display quality. The other type of software is a listing of a MATLAB program. In order to understand completely a mathematical algorithm or method, the authors recommend the user to program it himself, using MATLAB. In this way the user can also obtain all the intermediate results, thereby understanding what happens with a data set when analyzed by PCA. This section also contains an additional data set with which the user can experiment with the software. Some hints are given in to teachers and self-learners on how to use the material to optimize results.

1994