Machine Learning Bias, Statistical Bias, and Statistical Variance of Decision Tree Algorithms

By Thomas G. Dietterich, Eun Bae. 2008

203 citations recorded by the source corpus.

How researchers see this paper

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

In Evaluation & Benchmarks

Citations in the corpus

Ethical and social risks of harm from Language Models (2021, background): “In classical statistics, “bias” designates the difference between a model’s prediction and the ground truth (Dietterich and Kong, 1995); in machine learning, minimising statistical bias is a component of reducing error (Dietterich and Kong, 1995).”
Taxonomy of Risks posed by Language Models (2022, background): “In classical statistics, “bias” des-ignates the difference between a model’s prediction and the ground truth [47]; in machine learning, minimising statistical bias is a component of reducing error [47].”
On the asymptotics of random forests (2014, background): “For instance, trees can be built by randomizing the set of features (Dietterich and Kong, 1995; Ho, 1998), the data set (Breiman, 1996, 2000), or both at the same time (Breiman, 2001; Cutler and Zhao, 2001).”
Extremely randomized trees (2006, methodology): “…the statistical notions of variance and its companion, the bias, were studied more systematically by machine learning researchers (see for example, Dietterich and Kong, 1995; Breiman, 1996a; Friedman, 1997), and the high variance of tree-based models, like those induced by CART or C4.5, was…”
Understanding Random Forests: From Theory to Practice (2014, background): “], Dietterich and Kong [1995 ] propose to randomize the choice of the best split at a given node by selecting uniformly at random one of the 20 best splits of node t. The authors empirically show in [Dietterich and Kong, 1995 ; Dietterich, 2000 a] that randomizing in this way gives results that are ”