Figure 9

Graphical illustration of Principal Component Analysis in the case of a 3-dimensional representation space. Each dot shows the values of the variables pertaining to a given example. Left: all dots fall on a straight line, which means that each example can be described unambiguously by its abscissa along that line: the number of variables can be reduced can be reduced from 3 to 1 without information loss. Right: with negligible information loss, each example can be described by its coordinates with respect to the first two principal axes, so that the number of variables can be reduced from 3 to 2.