Introduction to Support Vector Machines

By Dustin Boswell. 2002

2,799 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/ea2ea7c6e280c1cfb67ee38ea63a327b1ba3ca36

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

A Sensor-Fusion Drivable-Region and Lane-Detection System for Autonomous Vehicle Navigation in Challenging Road Scenarios (2014, background): “New examples are then mapped into that same space and are predicted to belong to a category based on which side of the plane the examples will fall [53].”
Text-Independent Writer Identification and Verification Using Textural and Allographic Features (2007, methodology): “We mention that we replaced the linear distance combiner with an SVM [61], [62] trained for writer verification.”
Confusion Matrix Stability Bounds for Multiclass Classification (2012, methodology): “KHS). 4.1 Hilbert Space Regularized Algorithms Many well-known and widely-used algorithms feature a minimization of a regularized objective functions [9]. In the context of multiclass kernel machines [10,11], this regularizer Ω(h)may take the following form: Ω(h)= X q khqk2 k. where k :X ×X → Rdenot”
Learning with Kernels: Support Vector Machines, Regularization, Optimization, and Beyond (2003, methodology): “Gaussian process models have been used extensively in geostatistics (e.g., Cressie 1993)and also used in machine-learning contexts (e.g., Williams and Rasmussen 1996).”
Learning Halfspaces with Malicious Noise (2009, background): “In addition to its practical relevance, the problem has connections to many well-studied topics such as kernel methods [26], crypt ographic hardness of learning [15], hardness of approximation [6, 9], learning Boolea n circuits [2], and additive/multiplicative update learning algorithms [17, 7].”
A Novel Chemometric Method for the Prediction of Human Oral Bioavailability (2012, methodology): “Compared with traditional regression and neural networks methods, SVRs have some advantages, including global optimum, good generalization ability, simple implementation, few free parameters, and dimensional independence [40,41].”
Credit rating analysis with support vector machines and neural networks: a market comparative study (2004, unknown): “[7] [31].”
Variable Selection Using SVM-based Criteria (2003, background): “The support vector machine classiﬁer is a binary classiﬁer algorithm that looks for an optimal hyperplane as a decision function in a high-dimensional space (Boser et al., 1992, Vapnik, 1998, Cristianini and Shawe-Taylor, 2000).”
SVMTool: A general POS Tagger Generator Based on Support Vector Machines (2004, methodology): “This is achieved by working in the Support Vector Machines (SVM) learning framework (Cristianini and Shawe-Taylor, 2000), and by offering NLP researchers a highly customizable POS-tagger generator.”
Extracting Actionable Knowledge from Decision Trees (2007, background): “While decision-making and optimization techniques are not new, their application to data mining has resulted in a new generation of learning algorithms, such as support vector machines [38], [11], manifold regularization for subspace learning [5].”
A GA-based feature selection and parameters optimizationfor support vector machines (2006, background): “These concepts can also be found in (Kecman, 2001; Schőlkopf and Smola, 2000; Cristianini and Shawe-Taylor, 2000).”
Kernel-Based Nonlinear Blind Source Separation (2003, unknown): “based learning, which has become a popular technique (Vapnik, 1995; Cristianini & Shawe-Taylor, 2000; Sch ¨ olkopf & Smola, 2002; Burges, 1998; Sch ¨ olkopf, Smola, & M ¨ uller, 1998; M ¨ uller, Mika, R¨ atsch, Tsuda, & Sch ¨ olkopf, 2001).”
A Statistical Approach to Texture Classification from Single Images (2004, methodology): “For instance, Support Vector Machines (Cristianini and Shawe-Taylor, 2000; Hayman et al., 2004; Kim et al., 2002; Schólkopf and Smola, 2002), and perhaps more appropriately Relevance Vector Machines (Tipping, 2001), are both capable of reducing the number of models while providing good generalisatio”
Gene Selection for Cancer Classification using Support Vector Machines (2002, methodology): “This is the case for example of the mean-squared-error classifier (Duda, 1973) with cost function J = x∈X w · x − y 2 and linear SVMs (Boser, 1992; Vapnik, 1998; Cristianini, 1999), which minimize J = (1/2)||w|| 2 , under constrains.”
Lagrangian Support Vector Machines (2001, background): “The dual to the standard quadratic linear SVM (2) [16, 28, 17, 6] is the following: The variables ( w, γ ) of the primal problem which determine the separating surface (4) can be obtained from the solution of the dual problem above [19, Eqns.”
Value Regularization and Fenchel Duality (2007, methodology): “Very recently (while the present paper was under review), Dudı́k and Schapire (2006) used Fenchel duality to explore maximum entropy distribution estimation under constraints, and Altun and Smola (2006) explored relations between divergence minimization and statistical inference.”
Multicategory Support Vector Machines , Theory , and Application to the Classification of Microarray Data and Satellite Radiance Data (2004, methodology): “For an overview, see Vapnik (1998), Burges (1998), and Cristianini and Shawe-Taylor (2000). It is commonly known that the SVM paradigm can comfortably sit in regularization frameworks where we have a data fit component ensuring the model fidelity to data, and a penalty component enforcing the model ”
Sparse multinomial logistic regression: fast algorithms and generalization bounds (2005, methodology): “Thus, just like the margin bounds, the results we derive do not provide necessary conditions for good generalization: Even if kb w w w wk 1 is large (or, in the case of the SVM, if the margin separating the classes is small), good generalization may be possible.”
Impact of imputation of missing values on classification error for discrete data (2008, background): “To this end, this paper provides: • The largest number of popular and modern classifiers, namely, RIPPER [21] [22], C4.5 [17], KNN, support vector machine [23] [24], and Na¨ve-Bayes.”
Spam Filtering Using Inexact String Matching in Explicit Feature Space with On-Line Linear Classifiers (2006, background): “Following work on support vector machines (Boser et al., 1992) one may expect that providing the perceptron with higher margin will add to the stability and accuracy of the hypothesis produced (Cristianini and Shawe-Taylor, 2000).”