Nonlinear kernelled adaptive prediction method

Abstract

The invention discloses a nonlinear kernelled adaptive prediction method which comprises the following six steps: data preprocessing, subspace subdivision, subspace adaptive fitting control, subspace connection, new sample prediction and prediction output. The method is characterized in that after data is preprocessed, the integral space of the data is subdivided into a plurality of successive subspaces; the optimal kernels and parameters are adaptively selected on each subspace by using an intelligent sliding controller based on particle swarm support vector regression so as to form optimal local fitting hypersurfaces; then the optimal local fitting hypersurfaces are connected by using a three-point Lagrange interpolation method so as to form a final total-space regression prediction function; and finally, new data is predicted and output. By using the method disclosed by the invention, the characteristic space subdivision of data is realized, and the prediction speed for large-scale multi-attribute nonlinear data is improved; and the kernel functions adapting to the data distribution are screened adaptively by using the intelligent sliding controller based on particle swarm support vector regression, thereby optimizing integral parameters, and ensuring the precision and accuracy of fitting prediction.

Description

technical field [0001] The invention belongs to the field of artificial intelligence and relates to a nonlinear prediction method, in particular to a nonlinear kernelization adaptive prediction method, in particular to a fast and efficient kernel function-based nonlinear adaptive regression prediction method. Background technique [0002] The study of nonlinear predictive control is one of the core research contents of machine learning, and has become the focus of attention in academia and industry. The Support Vector Regression Machine (SVR), developed on the basis of statistical learning theory, transforms the nonlinear prediction problem into a kernel-based machine learning problem. Through the dot product of the kernel function, it is convenient to transform the nonlinear transformation into a linear problem in a high-dimensional space, which greatly simplifies the application calculation and the search of the nonlinear mapping function, thus solving many model analysis ...

AI Technical Summary

Problems solved by technology

This invention patent does not study the process of dynamic kernel function selection and parameter adjustment, nor the overall function fitting process

In addition, Ren Shuangqiao, Yang Degui and others published "Slice Support Vector Machine" in the Journal of Computer Science in January 2009, drawing on the basi

Images