A Short Introduction to Boosting

By Yoav Freund, Robert E. Schapire. 1999

How researchers see this paper

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

1. Dacheng Tao
2. Yoshua Bengio
3. Trevor Darrell
4. Luc Van Gool
5. Yi Yang
6. Xiaogang Wang
7. Bo Li
8. Mohit Bansal

Citations in the corpus

• Recent advances in convolutional neural networks (2015, background): “Object detection has been a long-standing and important problem in computer vision [114–116].”
• Estimating treatment effect heterogeneity in randomized program evaluation (2013, methodology): “The proposed methodology with the GCV statistic does not require cross-validation and hence is more computationally efficient than the commonly used methods for estimation of treatment effect heterogeneity such as Boosting [Freund and Schapire (1999), LeBlanc and Kooperberg (2010)], Bayesian additiv”
• FOSTER: Feature Boosting and Compression for Class-Incremental Learning (2022, methodology): “To achieve this, we design a new learning paradigm for class-incremental learning based on feature boosting [13] and compression [31].”
• Co-teaching: Robust training of deep neural networks with extremely noisy labels (2018, background): “However, it is commonly known that boosting and active learning are sensitive to outliers and noise, and a few wrongly selected instances can deteriorate the learning performance of the whole model [12, 3].”
• Surpassing Human-Level Face Verification Performance on LFW with GaussianFace (2014, methodology): “For this paper, we chose three popular representatives: SVM [12], logistic regression (LR) [17], and Adaboost [18].”
• Agnostically learning halfspaces (2005, unknown): “[13].”
• Disaggregation of component soil series on an Ohio County soil survey map using possibilistic decision trees (2014, background): “It could be useful to explore ensemblemethods such as bagging (Breiman, 1996), boosting (Freund and Schapire, 1999), stacking (Wolpert, 1992), Random Forests (Breiman, 2001; Chen et al.”
• Feature Weighting and Boosting for Few-Shot Segmentation (2019, background): “3 shows, our second contribution is a new boosted inference, motivated by the traditional ensemble learning methods which are robust to overfitting [9, 10].”
• Algorithmic Decision-Making Based on Machine Learning from Big Data: Can Transparency Restore Accountability? (2017, methodology): “Boosting , as invented by Robert Shapire in the 1990s, is a technique for optimizing a particular classifier (say, a decision tree) (Freund and Shapire 1999).”
• Learning under Concept Drift: an Overview (2010, background): “Ideas from boosting [52] are often employed, giving more attention to the instances which were misclassified.”
• Road crack detection using deep convolutional neural network (2016, methodology): “The Boosting method [22] composed of 100 weak classifiers with a maximum depth of 5 is trained via the OpenCV toolkit.”
• Boosting Variational Inference (2016, methodology): “The iterative updates are in the style of boosting or greedy error minimization [Freund and Schapire, 1995, Freund et al., 1999, Friedman et al., 2000, Li and Barron, 1999].”
• Privacy preserving data classification with rotation perturbation (2005, methodology): “The PerturBoost algorithm is basically an algorithm wrapping the AdaBoost algorithm [2] for processing the perturbed data.”
• A coarse-to-fine model for airport detection from remote sensing images using target-oriented visual saliency and CRF (2015, methodology): “al [13] employed the Adaboost learning algorithm [29] as a feature selector over a large set of texture features to select the best discriminative features, which is further used for detecting airports and identifying the textural characteristics of airports.”
• Boosted decision trees as an alternative to artificial neural networks for particle identification (2004, background): “Within the last few years a great improvement has been made[6, 7, 8].”
• Mapping the NFT revolution: market trends, trade networks, and visual features (2021, methodology): “We pick AdaBoost [40], an ensemble of weak learners (in our case, decision trees) whose output is combined into single score through a weighted sum.”
• Characterisation of Areal Surface Texture (2013, methodology): “Kogan Page Science, London Blunt L, Jiang X, Leach RK, Harris PM, Scott P (2008) The development of user-friendly software measurement standards for surface topography software assessment. Wear 264:389–393 Bruzzone AAG, Costa HL, Lonardo PM, Lucca DA (2008) Advances in engineering surfaces for funct”