Improved Support Vector Machine Oil Price Forecast Model Based on Genetic Algorithm Optimization Parameters

Abstract

An improved oil price forecast model that uses support vector machine (SVM) was developed. The new model, called the GA-SVM forecast model, is based on genetic algorithm (GA) optimization parameters. In traditional SVM models, penalty factor C and kernel function parameter σ are generally dependent on experience. These empirical parameters are difficult to accomplish the price data's change. Therefore, we used GA to optimize the parameter selection methods of SVM in accordance with training data, and improved SVM forecast precision. To verify the validity of the model, we selected and analyzed the Brent oil stock price data from 2001/12/27 to 2011/10/30. Data for 2009/07/30 to 2011/07/22 were designated as training data set, and those for 2011/08/08 to 2011/08/17 were employed for testing. Results show that the forecast efficiency of GA-SVM was better than that of traditional SVM.