REGRESSION TREES WITH UNBIASED VARIABLE SELECTION AND INTERACTION DETECTION

By Wei-Yin Loh. 2002

640 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/5447b71c96aaff6400ccf4267f373f184eec07d2

How researchers see this paper

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

1. Cynthia Rudin
2. Susan Athey
3. Wei-Yin Loh
4. Giles Hooker
5. R. Caruana
6. Achim Zeileis
7. Sarah Tan
8. Max Kuhn

Citations in the corpus

• Bayesian and Empirical Bayesian Forests (2015, background): “Loh (2002) and Hothorn et al. (2006) are example generalizations, both of which combat the multiple testing and other biases inherent in tree-building through a sequence of hypothesis tests.”
• Optimal trees for prediction and prescription (2018, methodology): “Another approach taken by many methods [34, 81, 63] is to separate the task of choosing the split variable from the task of choosing the split threshold.”
• Applied Predictive Modeling (2013, methodology): “Also, several techniques exist (Frank et al. 1998; Loh 2002; Chan and Loh 2004; Zeileis et al. 2008) that use more complex models in the terminal nodes, similar to M5 and Cubist.”
• Learning model trees from evolving data streams (2010, methodology): “Here fall RETIS (Karalic 1992), SUPPORT (Chaudhuri et al. 1994), SECRET (Dobra and Gherke 2002), GUIDE ( Loh 2002 ), SMOTI (Malerba et al. 2002) and LLRT (Vogel et al. 2007).”
• Partykit: a modular toolkit for recursive partytioning in R (2015, methodology): “The particularly influential algorithms include CART (classification and regression trees, Breiman et al., 1984), C4.5 (Quinlan, 1993), QUEST/GUIDE (Loh and Shih, 1997; Loh, 2002), and CTree (Hothorn et al., 2006) among many others (see Loh, 2014, for a recent overview).”
• Model-Based Recursive Partitioning (2008, methodology): “Unbiasedness: A desirable property which has been emphasized for many of the more recent tree algorithms—such as QUEST (Loh and Shih 1997), GUIDE (Loh 2002) or CTree (Hothorn et al.”
• A review on longitudinal data analysis with random forest (2022, unknown): “Examples include GUIDE [32] and conditional inference [23] approaches.”
• Fifty Years of Classification and Regression Trees (2014, methodology): “Chaudhuri & Loh (2002) and Loh (2006b) extended GUIDE to quantile and Poisson regression, respectively.”
• Bayesian Treed Models (2002, background): “Loh (2001) also considers linear regressions in terminal nodes, and attempts to correct for possible bias in selection of splits.”
• Regression trees for longitudinal and multiresponse data (2012, methodology): “If d= 1, these definitions reduce to those in the univariate GUIDE algorithm [Loh (2009), page 1716].”
• Gaining insight with recursive partitioning of generalized linear models (2013, methodology): “Furthermore, MOB provides functionality for many different type of node models that even exceeds the versatility of SUPPORT and GUIDE.”
• Accurate intelligible models with pairwise interactions (2013, methodology): “GUIDE tests pairwise interactions based on the χ 2 test [18].”
• Classification Trees With Unbiased Multiway Splits (2001, unknown): “Loh (2001) suggests a way to detect pairwise interactions among the variables in regression trees.”