Estimating a Kernel Fisher Discriminant in the Presence of Label Noise

By Neil D. Lawrence, Bernhard Schölkopf. 2001

227 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.

Citations in the corpus

Learning kernel logistic regression in the presence of class label noise (2014, methodology): “In our final set of controlled experiments, we compare rMKLR to state-of-theart nonlinear classifiers that have robustness properties: rKFD [24], the SVM, RP-Kernel-Perceptron [39], and the model-free method of StPMKL [46].”
Learning from Noisy Labels with Deep Neural Networks (2014, background): “1 Bottom-up Noise Model We assume that label noise is random conditioned on the true class, but independent of the input x (see [10] for more detail about this type of noise).”
Robust supervised classification with mixture models: Learning from data with uncertain labels (2009, methodology): “Indeed, the model of RLDA, proposed by Lawrence et al. [18] and its extension [19] are designed for only two classes.”
Confident Learning: Estimating Uncertainty in Dataset Labels (2019, background): “…used in the literature, including latent noise transition rates (Sukhbaatar et al., 2014; Goldberger and Ben-Reuven, 2017; Reed et al., 2015), latent prior of uncorrupted labels (Lawrence and Schölkopf, 2001; Graepel and Herbrich, 2001), and inverse noise rates (Katz-Samuels et al., 2017).”
Multiple Kernel Learning from Noisy Labels by Stochastic Programming (2012, background): “By assuming that the percentage of incorrectly labeled training examples is given, we approximate 1It is important to distinguish stochastic programming (Kall & Wallace, 1994) from stochastic optimization (Robbins & Monro, 1951).”
Using Deep Belief Nets to Learn Covariance Kernels for Gaussian Processes (2007, methodology): “For classification tasks, [11] model p(X) as a mixture ∑”
Robust Loss Functions under Label Noise for Deep Neural Networks (2017, methodology): “[14] proposed such a method, based on maximum-likelihood estimation of the model using the EM algorithm, in the context of ﬁsher linear discriminant classiﬁer.”
Learning with Marginalized Corrupted Features (2013, background): “A final interesting direction is to investigate the effect of marginalizing corrupted labels (Lawrence & Schölkopf, 2001).”
Robust binary hypothesis testing under contaminated likelihoods (2014, methodology): “Somewhat more related is the mixture modeling approach of [5], [6], which attempts to learn the co ntamination model using the EM algorithm.”
Classification in the presence of class noise using a probabilistic Kernel Fisher method (2007, unknown): “In particular, G-ellipse is analogous to the toy data set in Lawrence and Schölkopf [20].”
Kernel Methods and Machine Learning (2014, background): “Learning models of this type have been explored by numerous researchers [10, 63, 76, 147, 178, 179, 180, 181, 189, 226, 241].”
Label Noise-Tolerant Hidden Markov Models for Segmentation: Application to ECGs (2011, methodology): “This type of noise is the same as in previous experiments [1].”
Label-Noise Robust Logistic Regression and Its Applications (2012, background): “This problem was also reported in [13], However, the solution they resort to is simply to assume that there is a trusted validation set available.”
KPCA plus LDA: a complete kernel Fisher discriminant framework for feature extraction and recognition (2005, methodology): “Subsequent research saw the development of a series of KFD algorithms (see Baudat and Anouar [6], Roth and Steinhage [7], Mika et al. [8], [9], [10], Yang [11], Lu et al. [12], Xu et al. [13], Billings and Lee [14], Gestel et al. [15], Cawley and Talbot [16], and Lawrence and Schö lkopf [17]).”