The Research on Prediction of Chinese Petrolic Demand Based on Support Vector Machine

摘要

A forecasting model for Chinese petrolic demand using the regression algorithm of support vector machine was constructed. According to the selected embedding data volume, the forecasting model set up appropriate parameters in the toolbox of Libsvm in MATLAB2.10 and took parameters optimizations, and it used the data of Chinese petrolic consumption from 1990 to 2011 as the input vectors and output vectors. The radial basis function (RBF) is taken as the kernel function to establish the model. At the same time, three other kinds of curve fittings in the sense of least squares were made. The error analysis among these models showed that predictions using the established SVM model are reliable. At last, the trained support vector machine model is used to predict the petrolic demand from 2012 to 2015.