Interactive Decision Tree Learning and Decision Rule Extraction Based on the ImbTreeEntropy and ImbTreeAUC Packages

This paper presents two new R packages ImbTreeEntropy and ImbTreeAUC for building decision trees, including their interactive construction analysis, which is a highly regarded feature field experts who want to be involved in the learning process. functionality includes application of generalized entropy functions, such as Renyi, Tsallis, Sharma-Mittal, Sharma-Taneja Kapur, measure impurity node. provides non-standard measures choose an optimal split point attribute (as well splitting) by employing local, semi-global global AUC measures. The contribution both that thanks learning, user able construct tree from scratch or, if required, phase enables making regarding ambiguous situations, taking into account each its cut-off. main difference with existing solutions our provide mechanisms allow analyzing trees’ structures (several trees simultaneously) are built after growing and/or pruning. Both support cost-sensitive defining misclassification cost matrix, weight-sensitive learning. Additionally, structure model can represented rule-based model, along various quality measures, support, confidence, lift, conviction, addedValue, cosine, Jaccard Laplace.