You are working with the text-only light edition of "H.Lohninger: Teach/Me Data Analysis, Springer-Verlag, Berlin-New York-Tokyo, 1999. ISBN 3-540-14743-8". Click here for further information.
| You are working with the text-only light edition of "H.Lohninger: Teach/Me Data Analysis, Springer-Verlag, Berlin-New York-Tokyo, 1999. ISBN 3-540-14743-8". Click here for further information.
|
| See also: PCA |   |
Suppose you have a bottle of wine of unknown origin. You only know that it belongs to one of three kinds of wine: either a Grignolino, a Barbera, or a Barolo. In order to assign the correct origin you make a chemical analysis of the wine, determining 15 different variables, such as the concentration of ethanol, methanol, phenols, etc. In addition, a colleague of yours has analyzed 150 reference samples determining the same 15 variables. You can perform a PCA of the data and plot the unknown into the projection of the first two principal components. As shown in the figure below, the unknown sample falls into the region where the Grignolinos cluster. Thus you may well conclude that the unknown wine is a Grignolino.
You may now perform the following exercise to experience a similar example.
Last Update: 2004-Jul-03