"outliers?" The test of those that are farthest out can depend on
whether or not we include the marginal cases in the reference set.
It is the strong opinion of the author that automated rejection
of neural network training variable outliers should neuerhe
done.
The
recommended method is to plot a sample histogram. Careful scrutiny
and experienced judgment will invariably triumph over blind applica-
tion of statistics. The recently graduated, new researcher may be
surprised at how many times histograms similar to that shown in
Figure 16.6 will appear.
_J
H
H
h
M
mh
rfi-n
Fig.
16.6: Why we avoid automated outlier rejection
Whe
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month, and much more.