Class A has the property that its value for parameter x
is randomly distributed between 0 and 40, and class B has
the proerty that is value for parameter x is randomly distributed
between 0 and 50.
Remember that we are looking for nuggets
which explain part of the target problem, but not
necessarily all of it.
If we look at the above data with the target
question "class = A", it is clearly possible
to identify a sub-group of records with a value of x above
40. However, there is no similarly distinguishable sub-group
with the property "class <> A" or "class = B".
Although intuitively it seems that we should
be able to get the opposite answer by negating the problem,
it actually depends upon the data as to whether this will
indeed be the case.
