A New SVR Incremental Algorithm Based on Boundary Vector

Author

Xu Hongmin ; Wang Ruopeng ; Wang Kaiyi

Author_Institution

Dept. of Math. & Phys., Beijing Inst. of Petrochem. Technol., Beijing, China

fYear

2010

fDate

10-12 Dec. 2010

Firstpage

1

Lastpage

4

Abstract

In dealing with a large number of train samples, Support Vector Regression (SVR) algorithm is slow. In particular, while new sample is added, all the training samples must be re-trained. In this paper, a new SVR incremental algorithm is presented, which is based on boundary vector. The algorithm takes full advantages of the geometric information of training sample sets. The observed data of China´s GDP is used as a case study for the new algorithm. The computing results show that the new algorithm not only can guarantee the accuracy of machine learning and good generalization ability, but also can increase the learning speed of the algorithm than the classical SVR algorithm, and can be used rapid incremental learning.

Keywords

economic indicators; learning (artificial intelligence); regression analysis; set theory; support vector machines; vectors; China GDP; SVR incremental algorithm; boundary vector; geometric information; machine learning; rapid incremental learning; support vector regression algorithm; training sample sets; Algorithm design and analysis; Economic indicators; Forecasting; Machine learning; Prediction algorithms; Support vector machines; Training;

fLanguage

English

Publisher

ieee

Conference_Titel

Computational Intelligence and Software Engineering (CiSE), 2010 International Conference on

Conference_Location

Wuhan

Print_ISBN

978-1-4244-5391-7

Electronic_ISBN

978-1-4244-5392-4

Type

conf

DOI

10.1109/CISE.2010.5676955

Filename

5676955