RepTree
Rep Tree is a method to generate a decision tree from a given dataset. It is considered to be an extension of C45 by improving the pruning phase by using Reduced Error Pruning (REP). More information about REP can be found here.
The method uses a separate pruning dataset. For every subtree it checks whether the subtree could be replaced by a single node, without lowering the performance of the classifier on this pruning set. As such, the pruning method is simple, but often considered to be too agressive, ie. it might remove subtrees which are actually relevant.
Similar to C45, the generated decision trees are available to the users, can be interpreted and it can be motivated why the algorithm ordered the nodes in the tree in the given way. Hence, this method provides a high level of explainability.
Mentioned in:
This conceptual method was created by Alonso et al. (2019) to create natural explanations to children using single classifiers.
https://data-en-maatschappij.ai/tools/article-tool-taimReturn to the Introduction