Title :
Study on building a forecasting model with improved grey relational analysis and support vector machines and its application
Author :
Lin, Yao-jin ; Wu, Shun-xiang
Author_Institution :
Dept. of Comput. Sci. & Eng., Zhangzhou Normal Univ., Zhangzhou, China
Abstract :
In tradition, grey System treats any random variations as a variation in the grey value within a certain range, and the random process is treated as a time-varying grey process within a certain range. Grey System successfully utilizes accumulated generation data instead of original data to build forecasting model, which makes raw data stochastic weak, or reduces noise influence to a certain extent. However, only one factor has been considered in the conventional model. In most cases, prediction problems usually consist of more than one factor. Therefore, a grey relational analysis with Support Vector Machine (GASVM) is proposed in this study to deal with series problems with multi-factor. In this study, an admixture is presented based on Grey System and Support Vector Machines. Pretreatment modules which grey relational analysis attribution reduction algorithm course endow different weights to each influencing factor. In addition, the new influencing factors were regarded as input factors. Finally, the predicted performance is checked. The prediction results prove that this regression module helps to improve the prediction precision.
Keywords :
forecasting theory; grey systems; support vector machines; admixture; attribution reduction algorithm course; forecasting model; grey relational analysis; grey system; grey value; prediction precision; random process; support vector machines; time-varying grey process; Intelligent structures; Intelligent systems; Machine intelligence; Neural networks; Predictive models; Risk management; Support vector machine classification; Support vector machines; Training data; Upper bound;
Conference_Titel :
Grey Systems and Intelligent Services, 2009. GSIS 2009. IEEE International Conference on
Conference_Location :
Nanjing
Print_ISBN :
978-1-4244-4914-9
Electronic_ISBN :
978-1-4244-4916-3
DOI :
10.1109/GSIS.2009.5408352