Letter to the Editor: Stability of Random Forest importance measures

By Malu Luz Calle, Víctor Urrea. 2011

343 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Management Science and Operations Research

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 Management Science and Operations Research

Citations in the corpus

All Models are Wrong, but Many are Useful: Learning a Variable’s Importance by Studying an Entire Class of Prediction Models Simultaneously (2018, methodology): “This RF VI measure has been the topic of empirical studies (Archer and Kimes, 2008; Calle and Urrea, 2010; Wang et al., 2016), and several variations of the measure have been proposed (Strobl et al.”
An experimental study of the intrinsic stability of random forest variable importance measures (2016, background): “In the research of Calle and Urrea, MDG is more robust than MDA to small perturbations of the data [22].”
Overview of random forest methodology and practical guidance with emphasis on computational biology and bioinformatics (2012, background): “This may be related to the higher instability of the permutation VIM [13].”
A random forest based approach for predicting spreads in the primary catastrophe bond market (2020, background): “As an example coming from the area of bioinformatics, Calle & Urrea (2010)) showed that permutation importance rankings were unstable to small perturbations of a gene data set related to the prognosis bladder cancer.”
All Models are Wrong but many are Useful: Variable Importance for Black-Box, Proprietary, or Misspecified Prediction Models, using Model Class Reliance (2018, methodology): “This RF VI measure has been the topic of empirical studies (Archer and Kimes, 2008; Calle and Urrea, 2010; Wang et al., 2016), and several variations of the measure have been proposed (Strobl et al.”
The revival of the Gini importance? (2018, background): “On the other hand, for high dimensional data, the permutation importance is very computationally intensive and Calle and Urrea (2011) showed that”
Metabolomics and Multi-Omics Integration: A Survey of Computational Methods and Resources (2020, background): “Generally, metrics either evaluate the decrease in model fitness (e.g., Mean Decrease Accuracy [178] or Mean Decrease Gini [179,180]) when features are removed, or compare the influence of analytes on the model using the original values and shuffled values [181].”