Improved Boosting Algorithms Using Confidence-rated Predictions

By Robert E. Schapire, Yoram Singer. 1998

How researchers see this paper

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

Citations in the corpus

Boosting and Differential Privacy (2010, methodology): “Our algorithm is based on a variant of Freund and Schapire’s Adaboost algorithm [13], due to Schapire and Singer [24].”
Entity Resolution and Federated Learning get a Federated Resolution (2018, methodology): “One explanation comes from the fact that all learners, including "Ideal", use AdaBoost for the same number of iterations.”
Discriminative Reranking for Natural Language Parsing (2000, methodology): “As a final point, following Schapire and Singer (1998) we introduce some smoothing to prevent parameter estimates being undefined when either or =”
Predicting subscriber dissatisfaction and improving retention in the wireless telecommunications industry (2000, unknown): “Adaboost [3] was used for the decision tree, and a variant of Adaboost that is appropriate when a classifier produces confidence ratings [ 8 ], [9] was used for the neural network.”
Cost-sensitive boosting for classification of imbalanced data (2007, methodology): “In the past several years, many publications have reported that the AdaBoost algorithm has been successfully applied to most popular classifiers [25, 31, 81, 83].”
ML-KNN: A lazy learning approach to multi-label learning (2007, methodology): “MH algorithm proposed by Schapire and Singer [18] is employed to train the multi-label alternating decision trees.”
Boosting for Learning Multiple Classes with Imbalanced Class Distribution (2006, methodology): “The AdaBoost (Adaptive Boosting) algorithm [5] [19] is reported as an effective boosting algorithm to improve classification accuracies of any " weak " learning algorithms.”
Importance Sampled Learning Ensembles (2003, methodology): “At step m, the modified data distribution is given by (24) with”
Bregman Divergences and Surrogates for Learning (2009, background): “These losses play fundamental roles in some of the most popular supervised learning algorithms such as AdaBoost, [10], Additive logistic regression [5], Support Vector Machines [6].”
Boosting Neural Networks (2000, unknown): “Schapire & Singer (1998) addresses this issue.”
Model adaptation for spoken language understanding (2005, methodology): “This algorithm can be seen as a procedure for finding a linear combination of base classifiers which attempts to minimize an exponential loss function [9], which in this case is: 0 2 0 3 4 5 6 # 7 3 8 9 : ; # < 3 = >”
Round Robin Classification (2002, background): “Although the ﬁnal word on this issue remains to be spoken, it seems to be the case that techniques that include conﬁdence estimates into the computation of the ﬁnal predictions are preferable (cf. also Schapire and Singer, 1999).”