Reference : Closed-form dual perturb and combine for tree-based models

	Document type :	Scientific congresses and symposiums : Paper published in a book
	Discipline(s) :	Engineering, computing & technology : Computer science
	To cite this reference:	http://hdl.handle.net/2268/25764

Title :	Closed-form dual perturb and combine for tree-based models
Language :	English
Author, co-author :	Geurts, Pierre [Université de Liège - ULg > Dép. d'électric., électron. et informat. (Inst.Montefiore) > Systèmes et modélisation >]
	Wehenkel, Louis [Université de Liège - ULg > Dép. d'électric., électron. et informat. (Inst.Montefiore) > Systèmes et modélisation >]
Publication date :	2005
Main document title :	Proceedings of the International Conference on Machine Learning (ICML 2005)
On invitation :	No
Audience :	International
Event name :	22nd International Conference on Machine Learning
Event date :	2005
Event place (city) :	Bonn
Event country :	Germany
Keywords :	[en] machine learning ; optimisation
Abstract :	[en] This paper studies the aggregation of predictions made by tree-based models for several perturbed versions of the attribute vector of a test case. A closed-form approximation of this scheme combined with cross-validation to tune the level of perturbation is proposed. This yields soft-tree models in a parameter free way, and reserves their interpretability. Empirical evaluations, on classification and regression problems, show that accuracy and bias/variance tradeoff are improved significantly at the price of an acceptable computational overhead. The method is further compared and combined with tree bagging.
Permalink :	http://hdl.handle.net/2268/25764
Other URL :	http://www.montefiore.ulg.ac.be/services/stochastic/pubs/2005/GW05

	File(s) associated to this reference
	Fulltext file(s): File Commentary Version Size Access Open access geurts-icml-2005.pdf Publisher postprint 212.89 kB View/Open

All documents in ORBi are protected by a user license.

 

© ULg Library Network |  Feedback |  ORBi legal notices

Dspace v1.5