Training of support vector regressors based on the steepest ascent method

Author

Hirokawa, Y. ; Abe, Shigeo

Author_Institution

Graduate Sch. of Sci. & Technol., Kobe Univ., Japan

Volume

2

fYear

2002

fDate

18-22 Nov. 2002

Firstpage

552

Abstract

In this paper, we propose a new method for training support vector regressors. In our method, we partition all the variables into two sets: a working set that consists of more than two variables and a set in which variables are fixed. Then we optimize the variables in the working set using the steepest ascent method. If the Hessian matrix associated with the working set is not positive definite, we calculate corrections only for the independent variable in the working set. We test our method by two benchmark data sets, and show that by increasing the working set size, we can speed up training of support vector regressors.

Keywords

Hessian matrices; differential equations; function approximation; learning (artificial intelligence); optimisation; support vector machines; time series; Hessian matrix; Mackey-Glass differential equation; function approximation; optimization; steepest ascent method; support vector machines; support vector regressor training; time series; water purification plant; Benchmark testing; Differential equations; Function approximation; Lagrangian functions; Optimization methods; Purification; Quadratic programming; Support vector machine classification; Support vector machines; Training data;

fLanguage

English

Publisher

ieee

Conference_Titel

Neural Information Processing, 2002. ICONIP '02. Proceedings of the 9th International Conference on

Print_ISBN

981-04-7524-1

Type

conf

DOI

10.1109/ICONIP.2002.1198117

Filename

1198117