Second, it can help users to see whether nearest neighbor
is suitable for classification, or whether the distance defi-
nition needs modification such as using weighted distance
to reduce the interference of irrelevant attributes. Fig. 2 is
an example showing that nearest neighbor is not fit for the
Monks-2 problem: one of three Monks problems used in a
comparison of different learning techniques which was performed
at the 2nd European Summer School on Machine
Learning in 1991 [29]. The target function of the Monks-
2 problem is: EXACTLY TWO of {a1 = 1, a2 = 1, a3 =
1, a4 = 1, a5 = 1, a6 = 1}.