The following information was submitted:
Transactions: Please, select the Journal that you submit to
Transactions ID Number: 32-788
Full Name: Jiang Qiang
Position: Ph.D. Candidate
Age: ON
Sex: Male
Address: No.111, north section 1, second cycle road,chengdu,sichuan
Country: CHINA
Tel: +86 13908189345
Tel prefix:
Fax:
E-mail address: jq753428375@126.com
Other E-mails:
Title of the Paper: A kind of Nonlinear System States Prediction Based on Least Square Support Vector Machines
Authors as they appear in the Paper: jiang qiang, xiao jian, zheng gao
Email addresses of all the authors: jq753428375@126.com,jian_x@126.com,jq753428375@yahoo.cn
Number of paper pages: 11
Abstract: Nonlinear phenomena can be found in our life everywhere. Chaos is an important nonlinear phenomenon. Chaos prediction has played an important role in the study of chaos system. However, it is difficult that predict chaos, and previous methods have not perfect forecasting accuracy. A method is proposed to predict the states of chaos based on the algorithm of LS-SVM (least square support vectors machine) in this study. Our approach is based on reconstruct phase space coming from the Takens embedding theorem. In this approach, the data are divided into two parts; the first part data are used to train the model, another part data are used as the test set. The learning model can be obtained by moving the window, whose width is n, along the axis time. The n relates to the capacity of the input points, it has a best district. The relationship between the and the RMSE (root mean square error) is obtained by experiments. Theory analysis has also been done. Furthermore, the!
experiment with adding rand noise was done. The all results show that the method based on LS-SVM has better performance than neural network, it can be used effectively in chaos prediction.
Keywords: nonlinear; chaos; prediction; LS-SVM; phase space
EXTENSION of the file: .pdf
Special (Invited) Session:
Organizer of the Session:
How Did you learn about congress:
IP ADDRESS: 124.161.67.134
