TDIDT (Top-Down Induction of Decision Trees) Algorithm


Dataset  

Action to take when processing test data if there is a 'missing branch' of the decision tree
(as described in Section 6.7 of Principles of Data Mining)

use overall majority class (default)
leave as unclassified
use local majority class

Create log file of rules in CSV format?   no log file   create in type 1 format   create in type 2 format

trace1   yes no    Display messages relating to generation of decision tree (including clashes)
trace2   yes no    Display messages relating to processing of test data (including when there is a missing branch of the decision tree)