Greedy function approximation: A gradient boosting machine.

By Jerome H. Friedman. 2001

30,033 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Statistical and Nonlinear Physics, 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 Statistical and Nonlinear Physics

In Statistics and Probability

In Statistical Learning Theory

Citations in the corpus

Fully Decentralized Joint Learning of Personalized Models and Collaboration Graphs (2019, methodology): “We plan to extend our approach to (functional) gradient boosting (Friedman, 2001; Wang et al., 2015) where the graph regularization term would need to be applied to an inﬁnite set of base predictors.”
Visualizing the Feature Importance for Black Box Models (2018, background): “Methods for feature eﬀects include partial dependence (PD) plots [10], individual conditional expectation (ICE) plots [11] and, more recently, SHAP values [15].”
Explanations of model predictions with live and breakDown packages (2018, methodology): “An interesting and widely adopted tool for estimation of marginal model response is Partial Dependency Plot (Friedman, 2001), that presents the marginal relationship between the variable of interest and a single variable from the model.”
Visualizing the effects of predictor variables in black box supervised learning models (2016, methodology): “© 2020 Royal Statistical Society 1369–7412/20/821059 J. R. Statist. Soc. B (2020) 82, Part 4, pp.”
Making Tree Ensembles Interpretable (2016, background): “Ensemble models of decision trees such as random forests (Breiman, 2001) and boosted trees (Friedman, 2001) are popular machine learning models, especially for prediction tasks.”
Forest Floor Visualizations of Random Forests (2016, methodology): “See the grid of black and green lines on the model structure surface in Figure 1.”
Distilling Knowledge from Deep Networks with Applications to Healthcare Domain (2015, methodology): “Gradient boosting [15, 16] is a method which takes an ensemble of weak learners, usually decision trees, to optimize a differentiable loss function by stages.”
Interpreting tree ensembles with inTrees (2014, background): “Tree ensembles such as random forests [3] and boosted trees [7] are accurate supervised learners, and so are powerful in capturing information in the data.”
Interpretable classification models for recidivism prediction (2015, methodology): “tianini and Shawe-Taylor, 2000; Berk and Bleich, 2014). Stochastic Gradient Boosting: A popular black-box method that create prediction models in the form of an ensemble of weaker prediction models (Friedman, 2001; Freund and Schapire, 1997). 4.1.4. Details on Experimental Design, Parameter Tuning, ”
On the Fairness of Disentangled Representations (2019, methodology): “For the prediction, we consider the same gradient boosting classifier [27] as in [61] which was trained on 10 000 labeled examples (subsequently denoted by GBT10000) and which achieves higher accuracy than the crossvalidated logistic regression.”
Causal inference by using invariant prediction: identification and confidence intervals (2015, methodology): “For example, ˆ B could be the set of variables with non-zero regression coeﬃcient with square-root Lasso estimation [Belloni et al., 2011], Lasso [Tibshirani, 1996] or boosting [Schapire et al., 1998, Friedman, 2001, B¨uhlmann and Yu, 2003] with cross-validated penalty parameter.”
Mathematical optimization in classification and regression trees (2021, background): “Gradient Boosting Machines (GBMs) is also an iterative procedure based on combining weak tree learners but usually deeper than in AdaBoost.”
Fitted Q-Learning for Relational Domains (2020, methodology): “We employ Functional Gradient Boosting (Friedman 2001) in relational domains (Getoor 2007; De Raedt et al. 2016) to capture the Bellman error for every Q-learning iteration and obtain lifted basis functions in a non-parametric fashion.”
InterpretML: A Unified Framework for Machine Learning Interpretability (2019, background): “(Pedregosa et al., 2011), plotly (Plotly, 2015), lime (Ribeiro et al., 2016), shap (Lundberg and Lee, 2017), SALib (Herman and Usher, 2017), partial dependence (Friedman, 2001), Jupyter (Kluyver et al., 2016), pandas (McKinney), and more.”
Local Linear Forests (2018, methodology): “1 Methods In this section, we compare local linear forests, random forests, BART [Chipman, George, and McCulloch, 2010], and gradient boosting [Friedman, 2001].”