Extremely randomized trees

By Pierre Geurts, Damien Ernst, Louis Wehenkel. 2006

8,419 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Statistics and Probability, Statistical Learning Theory

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Statistics and Probability

In Statistical Learning Theory

Citations in the corpus

Findings of the 2014 Workshop on Statistical Machine Translation (2014, background): “A total of 221 features and the extremely randomised trees (Geurts et al., 2006) Different from the standard use of pseudo-reference features, these features do not assume that the alternative MT systems are better than the system of interest.”
Gradient-Based Neural DAG Learning (2019, methodology): “This is done by ﬁt-ting an extremely randomized trees (Geurts et al., 2006) (using ExtraTreesRegressor from scikit-learn ) for each variable against all the other variables.”
Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles (2016, background): “[16] later showed that bagging is unnecessary if additional randomness can be injected into the random subset selection procedure.”
Black-Box vs. White-Box: Understanding Their Advantages and Weaknesses From a Practical Point of View (2019, background): “decision tree; even, than other popular state-of-the-art classifiers, which are not based on decision trees [112]–[114].”
Generalized random forests (2016, background): “Our asymptotic theory relates to an extensive recent literature on the statistics of random forests, most of which focuses on the regression case (Arlot and Genuer, 2014; Biau, 2012; Biau et al., 2008; Biau and Scornet, 2016; Breiman, 2004; Bühlmann and Yu, 2002; Chipman et al., 2010; Denil et al., ”
Making AI Forget You: Data Deletion in Machine Learning (2019, background): “By quantizing or randomizing decision tree splitting criteria (such as in [36]) it seems possible to support efﬁcient deletion.”
Paraphrasing Revisited with Neural Machine Translation (2017, methodology): “To improve accuracy and control over-ﬁtting we built an ensemble of regression trees using the Extra-Trees algorithm (Geurts et al., 2006) which ﬁts a number of randomized decision trees (a.k.a. extra-trees) on various sub-samples of the dataset.”
Sequential Model-Based Ensemble Optimization (2014, methodology): “Random Forest We fix the number of trees to 100 and we explore two different ways of producing them: either the original Breiman [2001] method or the extremely randomized trees method of Geurts et al. [2006]. We also explore the choice of bootstrapping or not the training set before generating a tre”
Reinforcement Learning Strategies for Clinical Trials in Non-small Cell Lung Cancer (2011, methodology): “For instance, using random forest (RF) or extremely randomized trees (ERT) techniques is very effective for extracting well-fitted Q functions (Ernst, Geurts, and Wehenkel, 2005; Geurts, Ernst, and Wehenkel, 2006; Guez, Vincent, Avoli, and Pineau, 2008; Zhao et al., 2009).”
FBK-UPV-UEdin participation in the WMT14 Quality Estimation shared-task (2014, methodology): “3) with the features described in Section 2.1 using one learning algorithm: extremely randomized trees (ET) (Geurts et al., 2006).”
On the asymptotics of random forests (2014, methodology): “For instance, tree construction can be independent of D n (Biau, 2012), depend only on the X i ’s (Biau et al., 2008) or depend on the whole training set (Cutler and Zhao, 2001; Geurts et al., 2006; Zhu et al., 2012).”
Towards global empirical upscaling of FLUXNET eddy covariance observations: validation of a model tree ensemble approach using a biosphere model (2009, background): “Random splits are justiﬁed because the base search algorithm for the best split operates locally (at one node) only which has little meaning globally, i.e. for the performance of the entire tree (Geurts et al., 2006).”
Structured class-labels in random forests for semantic image labelling (2011, methodology): “We train the binary decision forest according to the extremely randomized trees algorithm [12].”
Exact Combinatorial Optimization with Graph Convolutional Neural Networks (2019, methodology): “rformance of full strong branching (FSB), our slow expert. We also compare against three machine learning branchers: the learning-to-score approach of Alvarez et al. [4] (TREES) based on an ExtraTrees[17] model, as well as the learning-to-rankapproachesfromKhalil etal.[29](SVMRANK) andHansknechtetal”
Referential translation machines for predicting semantic similarity (2015, methodology): “TREE is an ensemble learning method over randomized decision trees.”