Ismail, Shuhaida and Shabri, Ani and Samsudin, Ruhaidah (2011) A hybrid model of self-organizing maps (SOM) and least square support vector machine (LSSVM) for time-series forecasting. Expert Systems with Applications, 38 (8). pp. 10574-10578. ISSN 0957-4174
Full text not available from this repository.
Official URL: http://dx.doi.org/10.1016/j.eswa.2011.02.107
Supportvectormachine is a new tool from Artificial Intelligence (AI) field has been successfully applied for a wide variety of problem especially in time-seriesforecasting. In this paper, leastsquaresupportvectormachine (LSSVM) is an improved algorithm based on SVM, with the combination of self-organizingmaps(SOM) also known as SOM-LSSVM is proposed for time-seriesforecasting. The objective of this paper is to examine the flexibility of SOM-LSSVM by comparing it with a single LSSVMmodel. To assess the effectiveness of SOM-LSSVMmodel, two well-known datasets known as the Wolf yearly sunspot data and the Monthly unemployed young women data are used in this study. The experiment shows SOM-LSSVM outperforms the single LSSVMmodel based on the criteria of mean absolute error (MAE) and root mean square error (RMSE). It also indicates that SOM-LSSVM provides a promising alternative technique in time-seriesforecasting.
|Uncontrolled Keywords:||time series, least square support vector machine, self-organizing maps, forecasting|
|Deposited By:||Liza Porijo|
|Deposited On:||25 Oct 2012 07:00|
|Last Modified:||13 Feb 2017 02:34|
Repository Staff Only: item control page