Graphics Reference
In-Depth Information
9.2.1.3 Cut Point
Refers to a real value that divides the range into two adjacent intervals, being the
first interval less than or equal to the cut point and the second interval greater than
the cut point.
9.2.1.4 Arity
In discretization context, it refers to the number of partitions or intervals. For example,
if the arity of a feature after being discretized is
m
, then there will be
m
1 cut points.
We can associate a typical discretization as an univariate discretization. Although
this property will be reviewed in Sect.
9.3.1
, it is necessary to introduce it here for
the basic understanding of the basic discretization process. Univariate discretiza-
tion operates with one continuous feature at a time while multivariate discretization
considers multiple features simultaneously.
A typical discretization process generally consists of four steps (seen in Fig.
9.1
):
(1)
sorting
the continuous values of the feature to be discretized, (2)
evaluating
a cut
−
Fig. 9.1
Discretization process
