A Feature Selection Newton Method for Support Vector Machine Classification

By glenn. fung, Olvi L. Mangasarian. Computational optimization and applications, 2004

321 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/7edf1ae97a21f1f754078cac571df0341217e237

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

Residual Weighted Learning for Estimating Individualized Treatment Rules (2015, background): “(2003) and Fung and Mangasarian (2004) investigated the support vector machine (SVM) using thè1-norm penalty (Tibshirani 1994). Wang et al. (2008) applied the elastic-net penalty (Zou and Hastie 2005) for variable selection with an SVM-like method.”
Grafting: Fast, Incremental Feature Selection by Gradient Descent in Function Space (2003, background): “Jebara and Jaakkola (2000) also examine feature selection in a regularized risk minimization context for linear models. They define a prior describing likely values of weights, and choose one that tends to drive a number of weights to zero, in a similar fashion to l1 regularization. In fact l1 regul”
Supervised classification and mathematical optimization (2013, background): “It is common knowledge that the `1 norm yield, in general, sparse SVM classifiers, [100].”
Local Causal and Markov Blanket Induction for Causal Discovery and Feature Selection for Classification Part I: Algorithms and Empirical Evaluation (2010, methodology): “…on theL1-norm penalty (also known as lasso penalty), for example, feature selection via solution of theL1-norm formulation of SVM classification problem (Zhu et al., 2004; Fung and Mangasarian, 2004) and penalized least squares with lasso penalty on the regression coefficients(Tibshirani, 1996).”
A Comparison of Optimization Methods and Software for Large-scale L1-regularized Linear Classification (2010, methodology): “Readers interested in using L1-loss SVMs can refer to Bradley and Mangasarian (1998), Zhu et al. (2004), Fung and Mangasarian (2004), Mangasarian (2006), and references therein.”
Clustering Categories in Support Vector Machines (2017, background): “The most popular strategy proposed in the literature to achieve this goal is feature selection [12, 13, 26], which aims at selecting the subset of most relevant features for classi cation while maintaining or improving accuracy and preventing the risk of over tting.”
Binarized Support Vector Machines (2010, methodology): “We do not report results on the NLPSVM because outliers are not discussed in the paper of Fung and Mangasarian (2004). The classification ability of the linear SVM classifier dramatically worsens when introducing outliers and clearly underperforms the other two methods.”
MIXED-INTEGER SUPPORT VECTOR MACHINE (2009, methodology): “We used the LibSVM packages [21] and the L1SVM code from [22, 16].”
More generality in efficient multiple kernel learning (2009, methodology): “LP-SVM (Fung & Mangasarian, 2002) : This is the standard SVM formulation but with the l2 regularization on w replaced by l1 regularization.”
Variable selection for support vector machines via smoothing spline anova (2006, methodology): “Fung and Mangasarian (2004) developed a fast Newton algorithm to solve the dual problem of the 1-norm SVM. Another class of methods are the kernel scaling methods proposed by Weston, Mukherjee, Chapelle, Pontil, Poggio and Vapnik (2000) and Grandvalet and Canu (2002). A special issue on variable and”