Feature Induction based on Extremely Randomized Tree Paths

The volume of data generated and collected using modern technologies grows exponentially. This vast amount of data often follows a complex structure, significantly affecting the performance of various machine learning tasks. Despite the effort made, the problem of efficiently mining and analyzing such data is still persisting. Here, a novel data mining framework for unsupervised learning tasks is proposed based on decision tree learning and ensembles of trees. The proposed approach introduces an informative feature representation and is able to handle data diversity (e.g., numerical, canonical, etc.) and complexity (e.g., graphs, networks, data containing missing values etc.). Learning is performed in an unsupervised manner, following also the inductive setup. The experimental evaluation confirms the effectiveness of the proposed approach.