A Study on Sigmoid Kernels for SVM and the Training of non-PSD Kernels by SMO-type Methods

By Hsuan-Tien Lin. 2005

701 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/47e3fa7eeb0a88d64a9510c33ae256a35616edac

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

Pan evaporation prediction using a hybrid multilayer perceptron-firefly algorithm (MLP-FFA) model: case study in North Iran (2018, methodology): “Prior to modeling, the entire data sets were normalized to be between 0 and 1 to avoid issues related to larger numeric attributes dominating the importance of smaller numeric ranges in the MLP and SVM model (Lin and Lin 2003).”
An up-to-date comparison of state-of-the-art classification algorithms (2017, background): “The study in Lin and Lin (2003) proves that the sigmoid kernel is not better than the RBF kernel in general; moreover, the quality of the local minimum solution in parameters of the sigmoid kernel may not be guaranteed, it is therefore hard to select suitable parameters for the sigmoid kernel.”
Learning with Distance Substitution Kernels (2004, background): “Numerics: Convergence of SVM implementations to a (possibly local) stationary point can be guaranteed [9].”
A GA-based feature selection and parameters optimizationfor support vector machines (2006, methodology): “To implement our proposed approach, this research used the RBF kernel function for the SVM classifier because the RBF kernel function can analysis higher-dimensional data and requires that only two parameters, C and g be defined (Hsu, Chang, & Lin, 2003; Lin and Lin, 2003).”
A Practical Guide to Support Vector Classification (2009, background): “In addition, the sigmoid kernel behaves like RBF for certain parameters (Lin and Lin, 2003).”
Particle swarm optimization for parameter determination and feature selection of support vector machines (2008, background): “Lin et al. / Expert Systems with Applications 35 (2008) 1817–1824”
Support vector machine classification with indefinite kernels (2007, background): “Yet another option is to reformulate either the maximum marg in problem or its dual in order to use the indefinite kernel in a convex optimization problem (s ee [13]).”
Classification of EMG signals using PSO optimized SVM for diagnosis of neuromuscular disorders (2013, methodology): “To implement our proposed approach, the RBF kernel function is used for the SVM classifier since the RBF kernel function can analyze higher dimensional data [16,41].”
LIBSVM: A library for support vector machines (2011, background): “In (Lin and Lin, 2003), we have studied this issue in detail and proposed the following modification:”
Support vector machines in DSC‐based glioma imaging: Suggestions for optimal characterization (2010, background): “k, W could be considered as a scaling and shifting parameter of the tanh function (33).”
A Study on SMO-Type Decomposition Methods for Support Vector Machines (2006, unknown): “The above discussion is related to [19], which examines how t o solve (2) when it is non-convex.”