An Experimental Comparison of Genetic Algorithms for Optimizing Support Vector Regression in Grid Resources Prediction

摘要：

In order to manage the grid resources more effectively,the prediction information of grid resources is necessary in the grid system, In this study, support vector regression (SVR) is applied to grid resources prediction, In order to build an effective SVR model, SVRs parameters must be selected carefully. Therefore, we develop a genetic algorithmbased SVR (GA-SVR) model that can automatically determine the optimal parameters of SVR with higher predictive accuracy and generalization ability simultaneously. In this study, binary-coded GA (BGA), real-coded GA with heuristic crossover and uniform mutation (HRGA) and real-coded GA with simulated binary crossover and polynomial mutation (SRGA) are compared for SVRs parameters optimization with grid resources benchmark data set. Experimental results demonstrate that the hybrid model SRGA-SVR works better than BGA-SVR and HRGA SVR.

关键词： grid resources prediction support vector regression genetic algorithms

作者: Liang Hu Hongwei Li Guosheng Hu Xilong Che Hang Yang Weiqi Fan

作者单位: College of Computer Science and Technology, Jilin University ChangChun, China College of Mathematics, Jilin University ChangChun, China China Patent Information Center Beijing, China

会议类型: 国际会议

会议名称: The 10th International Conference on Intelligent Technologies(第十届智慧科技国际会议 InTech09)

会议地点: 桂林

会议语种:英文

页码: 664-670

在线出版日期: 2009-12-12（万方平台首次上网日期，不代表论文的发表时间）