Reference : Pruning randomized trees with L1-norm regularization
Scientific congresses and symposiums : Poster
Engineering, computing & technology : Computer science
Engineering, computing & technology : Electrical & electronics engineering
http://hdl.handle.net/2268/107193
Pruning randomized trees with L1-norm regularization
English
Joly, Arnaud mailto [Université de Liège - ULg > Dép. d'électric., électron. et informat. (Inst.Montefiore) > Systèmes et modélisation >]
Schnitzler, François mailto [Université de Liège - ULg > Dép. d'électric., électron. et informat. (Inst.Montefiore) > Systèmes et modélisation >]
Geurts, Pierre mailto [Université de Liège - ULg > Dép. d'électric., électron. et informat. (Inst.Montefiore) > Systèmes et modélisation >]
Wehenkel, Louis mailto [Université de Liège - ULg > Dép. d'électric., électron. et informat. (Inst.Montefiore) > Systèmes et modélisation >]
29-Nov-2011
A0
No
No
National
DYSCO Study Day
November 29, 2011
Leuven-Heverlee
Belgium
[en] Ensemble of randomized trees ; Pruning ; L1-norm regularization ; LASSO ; Supervised Learning ; Machine Learning
[en] Growing amount of high dimensional data requires robust analysis techniques. Tree-based ensemble methods provide such accurate supervised learning models. However, the model complexity can become utterly huge depending on the dimension of the dataset. Here we propose a method to compress such ensemble using random tree induced space and L1-norm regularisation. This leads to a drastic pruning, preserving or improving the model accuracy. Moreover, our approach increases robustness with respect to the selection of complexity parameters.
Systems and Modeling research unit
F. Schnitzler is supported by a F.R.I.A. scholarship. This work was also funded by the Biomagnet IUAP network of the Belgian Science Policy Office and the Pascal2 network of excellence of the EC. The scientific responsibility is the authors'.
http://hdl.handle.net/2268/107193

File(s) associated to this reference

Fulltext file(s):

FileCommentaryVersionSizeAccess
Open access
poster.pdfPublisher postprint1.57 MBView/Open

Additional material(s):

File Commentary Size Access
Private access
poster.svg767.87 kBRequest copy

Bookmark and Share SFX Query

All documents in ORBi are protected by a user license.