Abstract:The support vector machine (SVM) has recently been successfully used for classification and regression problems. As the core of SVM, the kernel function is used to map the input vectors into a higher dimensional feature space to realize the non-linear algorithm for original space. The selection of kernel function and corresponding parameters directly affects the learning ability and generalization ability of the support vector regression (SVR). This paper compares the performance of different kernel functions on the energy efficiency data obtained from the UCI repository. The kernel functions used are as follows: Linear, Gaussian, Polynomial, Sigmoid, Inv. Multiquadric and the Semi Local kernel. The influence of parameters on the mean square error of cross validation (MSECV) for each kernel is shown in the paper. The combination of SVR parameters and kernel parameter with minimum MSECV obtained by grid search method based on cross-validation is chosen for SVR model building. The experiment results based on the dataset showed that the Semi Local kernel has the best predictive capability. The mean-squared error of prediction (MSEP) of the Semi Local kernel was 97.13%, 8.26%, 0.96%, 99.95%, 27.36% lower than the value of the Linear, Gaussian, Polynomial, Sigmoid and Inv. Multiquadric kernel.

A Rapid Parameter Selection Method of Kernel Function.

SVM KERNEL PARAMETER SELECTION BASED ON IMPROVED GSO

Novel Linear Search for Support Vector Machine Parameter Selection

The Optimization for Parameter of Kernel Function with GA

A Kernel Function Optimization And Selection Algorithm Based On Cost Function Maximization

Kernel Parameter Selection for Support Vector Machine Classification

Cross-validatory Framework for Optimal Parameter Estimation of KPCA and KPLS Models

SVM Parameter Selection Algorithm Based on Maximum Kernel Similarity Diversity

A Gradient-based Kernel Optimization Approach for Parabolic Distributed Parameter Control Systems

SVR Kernel Parameters Selection Based on Steady-State Genetic Algorithm

An Iterative Algorithm for Selecting the Parameters in Kernel Methods.

Rapid and brief communication: Learning the kernel parameters in kernel minimum distance classifier

Super-parameter Selection for Gaussian-Kernel SVM Based on Outlier-Resisting

Efficient Model Selection for Support Vector Machine with Gaussian Kernel Function

Performance evaluation of kernel functions based on grid search for support vector regression

Optimal Gaussian Kernel Parameter Selection for SVM Classifier

An Analytical Approach to Fast Parameter Selection of Gaussian Rbf Kernel for Support Vector Machine

Hyper Parameters Selection For Svr Based On Steady-State Genetic Algorithm

Parameter Selection for Support Vector Machines Based on Hybrid Genetic Algorithms

Data-Dependent Kernel Function Based Kernel Optimization Algorithm

Fast model selection by limiting SVM training times