Multi-class AdaBoost ∗

By Trevor J. Hastie, Saharon Rosset, Ji Zhu, Hui Zou. 2009

1,972 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 Artificial Intelligence

Citations in the corpus

Explaining Predictions from Tree-based Boosting Ensembles (2019, methodology): “y solution to our problem, we take the methodfrom[13],explainitandarticulatehowitcouldbeextended to accommodate tree-based boosting ensembles. In this extended abstract, we focus on adaptive boosting [6] •rst to disentangle the sequential training nature of boosting methods. 3 A CASE STUDY IN EXPLAI”
Multi-Class Deep Boosting (2014, methodology): “This includes the SAMME algorithm [Zhu et al., 2009] as well as the algorithm of Mukherjee and Schapire [2013], which is shown to be, in a certain sense, optimal for this approach.”
Snuba: Automating Weak Supervision to Label Training Data (2018, methodology): “We compare to pruned decision tree [43] and boosting [17] (AdaBoost), which use the labeled dataset to generate one complex or multiple, simple decision trees, respectively.”
Beyond sigmoids: How to obtain well-calibrated probabilities from binary classifiers with beta calibration (2017, methodology): “The second is the one implemented in Python’s Scikit-learn, based on Adaboost method SAMME (Zhu et al., 2009), referred to as Ada-S.”
IPBoost - Non-Convex Boosting via Integer Programming (2020, methodology): “R error function as used in the (multi-class) AdaBoost variant in [37].”
Learning with Marginalized Corrupted Features (2013, positive): “The derivation above can readily be extended to multiclass exponential loss (Zhu et al., 2006) (with K classes) by replacing the weight vector w by a D×K weight matrix W, and by replacing the labels y by label vectors y = {1,− 1 K−1} with ∑K k=1 yk = 0.”
User Classification with Multiple Textual Perspectives (2016, background): “ Adaboost: a popular ensemble learning approach proposed by (Zhu et al., 2009).”
Unifying Heterogeneous Classifiers With Distillation (2019, methodology): “These techniques have been successfully used in many applications, e.g ., multi-class clas-siﬁcation [15], object detection [34, 27], tracking [1], etc .”