A Walk-through-guide for Using Decision Trees Software - Data Mining with Decision Trees: Theory and Applications - page 154

Fig. 10.2 Data view in Weka.
To train the J48 classifier, first go classifier tab (the second tab right to
the preprocess tab). Figure 10.3 presents the resulted user interface. In this
window, you can select the learning algorithm (with the Choose button),
set the algorithm's parameters, select the evaluation procedure (the default
is 10 folds cross-validation) and change the target attribute (by default the
last attribute is considered as the target attribute).
Click the choose button and select the J48 algorithm under the trees
folder as illustrated in Figure 10.4.
It is possible to set the algorithm's parameters by clicking the white
text box that contains the algorithm's title along with its current list of
parameters' values. Clicking this box opens up a window that lists all the
parameters of J48 algorithm that can be set by the user, including:
(1) minNumObj — A numeric parameter that indicates the minimum
number of instances that a leaf should have.
(2) binarySplits — A binary parameter that indicates whether nominal
attributes should be partitioned using binary splits.