Smooth And Consistent Probabilistic Regression Trees - Archive ouverte HAL
Conference Papers Year : 2020

Smooth And Consistent Probabilistic Regression Trees

Abstract

We propose here a generalization of regression trees, referred to as Probabilistic Regression (PR) trees, that adapt to the smoothness of the prediction function relating input and output variables while preserving the interpretability of the prediction and being robust to noise. In PR trees, an observation is associated to all regions of a tree through a probability distribution that reflects how far the observation is to a region. We show that such trees are consistent, meaning that their error tends to 0 when the sample size tends to infinity, a property that has not been established for similar, previous proposals as Soft trees and Smooth Transition Regression trees. We further explain how PR trees can be used in different ensemble methods, namely Random Forests and Gradient Boosted Trees. Lastly, we assess their performance through extensive experiments that illustrate their benefits in terms of performance, interpretability and robustness to noise.
Fichier principal
Vignette du fichier
NeurIPS-2020-smooth-and-consistent-probabilistic-regression-trees-Paper.pdf (658.73 Ko) Télécharger le fichier
Origin Files produced by the author(s)

Dates and versions

hal-03050168 , version 1 (10-12-2020)

Identifiers

  • HAL Id : hal-03050168 , version 1

Cite

Sami Alkhoury, Emilie Devijver, Marianne Clausel, Myriam Tami, Éric Gaussier, et al.. Smooth And Consistent Probabilistic Regression Trees. NeurIPS 2020 - 34th International Conference on Neural Information Processing Systems, Dec 2020, Virtuelle, France. pp.1-11. ⟨hal-03050168⟩
195 View
480 Download

Share

More