(1*) C4.5 Syntax(用户自定义函数, 并非matlab自带函数):
predicted_targets = C4_5(training_patterns,training_targets, test_patterns, input parameter);
C4.5的matlab函数源代码, 可以参考Chapter 8 (记录在: Computer Manual in MatLab to accompany Pattern Classification工具箱的说明手册, 与 https://bbs.pinggu.org/thread-1087526-1-1.html),

(2) 画树图的指令可以用: dendrogram 或者 treeplot (语法可参考:
dendrogram: http://www.mathworks.com/help/stats/dendrogram.html#btk2sv9)
treeplot        :  http://www.mathworks.com/help/matlab/ref/treeplot.html?searchHighlight=treeplot

* Remark:
Function name: C4_5
Description:
Construct a decision tree recursively so as tominimize the error on a training set. Discrete features are split using
a histogram and continuous features are splitusing an information criteria. The algorithm is implemented under
the assumption that a pattern vector with fewerthan 10 unique values is discrete, and will be treated as such.
Other vectors are treated as continuous. Note thatdue to MATLAB memory and processing restrictions, the
recursion depth may be reached during theprocessing of a large complicated data set, which will result in an
error.