ESPCI web site
Decision Tree for uncertainty measures
The ensemble methods are popular machine learning techniques which are powerful when one wants to deal with both classification or prediction problems. A set of classifiers is constructed, and the classification or the prediction of a new data instance is done by tacking a weighted vote. These classifiers can be regression or classification trees. Practically, a tree is a piece-wise constant estimator on partitions obtained from the data. These partitions are induced by recursive dyadic split of the set of input variables. For example, CART (Classification And Regression Trees) is an algorithm which constructs a tree. The goal is to partition the space of input variable values in the most "homogeneous" K disjoint regions possible. More precisely, each partitioning value has to minimize a risk function.
However, in practice, experimental measures can be observed with uncertainty. This work proposes to extend CART algorithm to these kind of data. We present an induced model adapted to uncertainty data and both a prediction and split rule for a tree construction taking into account the uncertainty of each quantitative observation from the data base.
Joint work with Sami ALKHOURY, Myriam TAMI, Emilie DEVIJVER, Eric GAUSSIER