An Introduction to Support Vector Machines and Other Kernel-Based Learning Methods

By Tong Zhang. 2001

9,108 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 Statistical Learning Theory

Citations in the corpus

A Survey of Available Corpora For Building Data-Driven Dialogue Systems: The Journal Version (2015, methodology): “Given a user utterance, their system generated a response by following a two-step procedure: first, a sequence of topic keywords, used to create a seed reply, was extracted from the user’s utterance; second, starting from the seed reply, two separate Markov chains generated the words preceding and…”
Opening black box Data Mining models using Sensitivity Analysis (2011, methodology): “This includes a wide range of methods, such as neural networks (NN) (e.g. multilayer perceptrons and radial basis-functions) [4], Support Vector Machines (SVM) and other kernel-based methods [5], and even ensembles (where multiple models are combined to achieve a better predictive performance) [6].”
Interpretable classification models for recidivism prediction (2015, methodology): “The Radial Basis Function (RBF) kernel allows the method to handle classiﬁcation problems where the decision-boundary may be non-linear (see, e.g., Cristianini and Shawe-Taylor 2000, Berk and Bleich 2014).”
On the Schoenberg Transformations in Data Analysis: Theory and Illustrations (2010, neutral): “3 Some properties of the Schoenberg transformations”
Kernel Continual Learning (2021, methodology): “Kernel methods (Sch ¨ olkopf & Smola, 2002) provide a well-suited tool for this due to their non-parametric nature, and have proven to be a powerful technique in machine learning (Cristianini et al., 2000; Smola & Sch ¨ olkopf, 2004; Rahimi & Recht, 2007; Sinha & Duchi, 2016).”
Comprehensible credit scoring models using rule extraction from support vector machines (2007, neutral): “…i ∈ IR n and corresponding binary class labels y i ∈ {− 1 , +1 , the SVM classiﬁer, according to Vapnik’s original formulation satisﬁes the following conditions [8, 25]: which is equivalent to The non-linear function ϕ ( ) maps the input space to a high (possibly inﬁnite) dimensional feature space.”
Kernel Methods for Relation Extraction (2002, methodology): “The normalization corresponds to the standard unit norm normalization of examples in the feature space corresponding to the kernel space (Cristianini and Shawe-Taylor, 2000): p For both SVM Light and Voted Perceptron, we used their standard conﬁgurations (e.g., we did not optimize the value of C…”
Computational Methods in Drug Discovery (2014, methodology): “This strategy, however, must be offset by the fact that higher dimensional space creates more computational burden and contributes to overfitting (Cristianini and Shawe-Taylor, 2000).”
Class-specific material categorisation (2005, methodology): “More specifically, the goals in this section are (i) to study which descriptor is best suited to the task; (ii) to study the performance of SVMs relative to Nearest-Neighbour Classifiers (NNCs); and (iii) to investigate the improvement in performance as more samples are introduced into the training ”
Use of support vector regression and numerically predicted cloudiness to forecast power output of a photovoltaic power plant in Kitakyushu, Japan (2012, methodology): “According to Cristianini and Shawe‐Taylor [9], SVMs are based on recent advances in statistical learning theory and can deliver state‐of‐ the‐art performance in categorization and classification problems.”
Covering Number Bounds of Certain Regularized Linear Function Classes (2002, methodology): “Note that the corresponding bound given by Cristianini and Shawe-Taylor (2000) was their Theorem 4.19.”
Random k -Labelsets: An Ensemble Method for Multilabel Classification (2007, methodology): “The SVM was trained with a linear kernel and the complexity constant C equal to 1.”
Input Switched Affine Networks: An RNN Architecture Designed for Interpretability (2016, background): “Model classes whose decisions are naturally interpretable include logistic regression (Freedman, 2009), decision trees (Quinlan, 1987), and support vector machines with simple (e.g. linear) kernels (Andrew, 2013).”