Mapping Associations - Mapping Scientific Frontiers: The Quest for Knowledge Visualization - page 122

Fig. 3.32 Individual
differences scaling results of
two red-green color-deficient
subjects. The Y axis is not
fully extended as normal
subjects
Fig. 3.33 SCI weighted individual differences scaling display (Reproduced from Morris and
McCain 1998 )
PCA and MDS have been routinely used to reduce the dimensionality of linear
data. Euclidean distances provide reliable measures of a linear structure in a
high-dimensional space. The problem is that when we deal with a non-linear
structure, Euclidean distances may not be able to detect the true structure. The