C4.5 (Quinlan, 1992) builds a decision tree using the standard TDIDT (top-down induction of decision trees) approach, recursively partitioning the data into smaller subsets, based on the value of an attribute. At each step in the construction of the decision tree, C4.5 selects the attribute that maximizes the information gain ratio. The induced decision tree is pruned using pessimistic error estimation (Quinlan, 1992). There are several parameters that can be adjusted to alter the behavior of C4.5. In our experiments with C4.5, we used the default settings for all parameters. We used the C4.5 source code that is distributed with (Quinlan, 1992).