Trees, Probabilistic Functional

Decision Trees are one of most used models for prediction tasks in machine learning, data mining, and statistics. The representation language of a standard decision tree uses splitting tests based on a single input variable and constants at tree leaves. The most recent advances in learning decision trees from data involve the use of splitting tests based on combinations of the input variables and functions at tree leaves. These types of models are denominated functional trees. In this article, we present the issues and actual trends in learning functional trees from data. Keywords: functional trees; decision trees; discriminant analysis; multivariate data analysis; machine learning