Context-dependent feature analysis with random forests

Antonio Sutera, Gilles Louppe, Vân Anh Huynh-Thu, Louis Wehenkel, Pierre Geurts
Proceedings of the 32nd Conference on Uncertainty in Artificial Intelligence, PMLR R14:702-711, 2016.

Abstract

For many problems, feature selection is often more complicated thanidentifying a single subset of input variables that would togetherexplain the output. There may be interactions that depend oncontextual information, i.e., variables that reveal to be relevantonly in some specific circumstances. In this setting, the contributionof this paper is to extend the random forest variable importancesframework in order (i) to identify variables whose relevance iscontext-dependent and (ii) to characterize as precisely as possiblethe effect of contextual information on these variables.The usage and the relevance of our framework for highlighting context-dependent variablesis illustrated on both artificial and real datasets.

Cite this Paper

BibTeX
@InProceedings{pmlr-vR14-sutera16a, title = {Context-dependent feature analysis with random forests}, author = {Sutera, Antonio and Louppe, Gilles and Huynh-Thu, V{\^a}n Anh and Wehenkel, Louis and Geurts, Pierre}, booktitle = {Proceedings of the 32nd Conference on Uncertainty in Artificial Intelligence}, pages = {702--711}, year = {2016}, editor = {Ihler, Alexander and Janzing, Dominik}, volume = {R14}, series = {Proceedings of Machine Learning Research}, month = {25--29 Jun}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/r14/main/assets/sutera16a/sutera16a.pdf}, url = {https://proceedings.mlr.press/r14/sutera16a.html}, abstract = {For many problems, feature selection is often more complicated thanidentifying a single subset of input variables that would togetherexplain the output. There may be interactions that depend oncontextual information, i.e., variables that reveal to be relevantonly in some specific circumstances. In this setting, the contributionof this paper is to extend the random forest variable importancesframework in order (i) to identify variables whose relevance iscontext-dependent and (ii) to characterize as precisely as possiblethe effect of contextual information on these variables.The usage and the relevance of our framework for highlighting context-dependent variablesis illustrated on both artificial and real datasets.}, note = {Reissued by PMLR on 04 October 2026.} }
Endnote
%0 Conference Paper %T Context-dependent feature analysis with random forests %A Antonio Sutera %A Gilles Louppe %A Vân Anh Huynh-Thu %A Louis Wehenkel %A Pierre Geurts %B Proceedings of the 32nd Conference on Uncertainty in Artificial Intelligence %C Proceedings of Machine Learning Research %D 2016 %E Alexander Ihler %E Dominik Janzing %F pmlr-vR14-sutera16a %I PMLR %P 702--711 %U https://proceedings.mlr.press/r14/sutera16a.html %V R14 %X For many problems, feature selection is often more complicated thanidentifying a single subset of input variables that would togetherexplain the output. There may be interactions that depend oncontextual information, i.e., variables that reveal to be relevantonly in some specific circumstances. In this setting, the contributionof this paper is to extend the random forest variable importancesframework in order (i) to identify variables whose relevance iscontext-dependent and (ii) to characterize as precisely as possiblethe effect of contextual information on these variables.The usage and the relevance of our framework for highlighting context-dependent variablesis illustrated on both artificial and real datasets. %Z Reissued by PMLR on 04 October 2026.
APA
Sutera, A., Louppe, G., Huynh-Thu, V.A., Wehenkel, L. & Geurts, P.. (2016). Context-dependent feature analysis with random forests. Proceedings of the 32nd Conference on Uncertainty in Artificial Intelligence, in Proceedings of Machine Learning Research R14:702-711 Available from https://proceedings.mlr.press/r14/sutera16a.html. Reissued by PMLR on 04 October 2026.

Related Material