Graphics Reference
In-Depth Information
Figure
.
.
Variable and unit representations ater the deletion of the less important units
In exploratory MDA, to evaluate the influence of anomalous observations, statis-
ticians remove units having the greatest contributions and, projecting these as sup-
plementary points, compare the two-point configurations.
High-contribution unitsarethenmadesupplementary usingthedeletion tooland
the resulting planes are shown in Fig.
.
.
Itisevident that aremarkable change occurred:whilethe variables preserves their
position on thefirstaxis with respecttoFig.
.
,they moveonthe second factor,that
is, they are more influenced by anomalous observations.
he user can reiterate the point deletion step or, if the point configuration is be-
lieved to be su
ciently stable, go to the analysis of the classification step.
Figure
.
.
Units with a contribution
% (SHIW data)
