Smooth Boosting and Learning with Malicious Noise

By Rocco A. Servedio. 2001

213 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

Robust Estimators in High Dimensions without the Computational Intractability (2016, background): “A sequence of works studied this problem for the class of halfspaces [44], [45], [46] in the setting where the underlying distribution is a Gaussian, culminating in the work of Awasthi et al.”
Can machine learning be secure? (2006, background): “Specifically, they proved that if the learner is to perform correctly, in general the fraction of training points controlled by the adversary must be less than /(1+ ), where is the desired bound on classification errors by the learner [4, 6, 30].”
Learning Halfspaces with Malicious Noise (2009, methodology): “Lemma 7 ([25]).”
Adversarial attacks against intrusion detection systems: Taxonomy, solutions and open issues (2013, background): “Interesting solutions are based on boosting [135], and on the framework of robust statistics [131].”
Boosting and Hard-Core Set Construction (2003, methodology): “Building on this algorithm, Servedio has given an improved boosting-by-sampling algorithm which is also 1/� -smooth but uses only O(1/(�γ 2 )) stages ( Servedio, 2001 ).”
Mixture models, robustness, and sum of squares proofs (2017, background): “There has been a long line of work on making algorithms tolerant to error in supervised settings [41, 66], especially for learning halfspaces [4, 28, 42, 60], and for problems such as PCA [15, 16, 47, 70].”
The power of localization for efficiently learning linear separators with noise (2013, unknown): “[Ser01] Rocco A.”
Attribute-efficient learning of decision lists and linear threshold functions under unconcentrated distributions (2006, methodology): “SmoothBoost [27] has two parameters, γ and θ.”
Distribution-Specific Agnostic Boosting (2009, background): “In its least restricted version the examples are generated from some unknown distribution A over X ×{− 1 , 1 } .”
An Introduction to Boosting and Leveraging (2002, methodology): “The version of SmoothBoost described in [178] requires two parameters as input: (i) κ, which measures the desired error rate of the final classifier, (ii) and γ which measures the guaranteed edge of the hypothesis returned by the weak learner.”
Edited AdaBoost by weighted kNN (2010, methodology): “In the case of probabilistic setup, the expectation of loss [15] is”